The ultimate DIY? A Stax SRM-T2!

[jokerman777]

1 hour ago, justin said:

what about the 90C version which is 900v instead of 800v

thanks! Never thought of this obvious option, not so resourceful I am : )

[jamesmking]

update on the RIP T2.

The four 270K resistors R65 to R68 are discoloured I believe because of the higher than expected DC voltages across them from the massive DC imbalance. Otherwise there are no visible signs of problems.

I found a dead 2SK216 in addition to the dead 2SJ79 on the same channel that the 6922 failed on.

Diode testing found the dead J79 but to find the dead K216 I used my peak dca75 to measure the two ksa1156s and found radically different hfe measurements (43 vs 143) so I removed the other J79 and the readings were dd not change. The removed J79 measured ok. So I suspected the K216es and removed them 216es one measured good and one bad. The good K216 was connected to the bad J79 and good J79 connected to the bad K216... The ksa1156es measured the same hfe when the 4 transistors were removed so I believe they are ok. I'm replacing all 4 mosfets with ksa1220A and KSC2690A because that's all I have...

The other channel had the same imbalance issues and one of the K216es there is dead too.

Both channels now have 1220 and 2690a installed. But I noticed that the ksa1156s on the channel with the failed 6922 measured hfe 51 in circuit and the channel with only the single failed k216 the same ksa1156s measured 2hfe. Hmm...

looking at the circuit I decided to measure the resistance across led between the emitter and base of the ksa1156es... channel with the failed 6922 about 150K and about 190K leads revered. Channel with single bad k216 about 3.4M and open when reverse biased...

So either the led or the ksa1156s where leaking. Desoldered the ksa1156s and tested them... ok. desoldered the D23 led and replaced it. Now 3.4M and open circuit. So the led had gone leaky. At least both channels now measure the same... The other led close to the EL34 (D24
) on the channel with the dead 6922 is also leaky .

So score so far:

right channel

6922 shorted heater to cathode

K216 DCA75 measures it as a zener diode

J79 DCA75 measures it as two diodes

Led Brymen bm869s measures it as a 150K resistor

Led Brymen bm869s measures it as a 88K resistor

Left channel 

K216 DCA75 measures it as zener diode

 

Edited August 10, 2023 by jamesmking

[jamesmking]

The unofficial mostly modern T2 troubleshooting and verification guide.

Weclome, this guide covers the mostly modern T2 which has the following schematic:

The gerbers for the amp can be found here:

https://drive.google.com/drive/folders/1r3g2TAtBUaBdiMorTWX7yYgeJ7maQbYW

and the gerbers are staxt2nc3fdh7.zip

The guide is intended for both pre-power on verification of an amp build, verification of voltages on power on and general troubleshooting. All tests are performed using a Brymen BM869s and peak DCA75. Using a different multimeter may effect the results slightly but you should still get similar behaviours and ball-park figures.

Transistor and Diode Location

 

Section 1. Build verification with no valves installed and no wires/volume control connected to the amp

All of these checks can be performed with a multimeter from the top side of the amp board

 

leds Checks.

All 3 leds on each channel should read about 0.55V voltage drop in forward bias (multimeter  + terminal/red lead connected to led + and multimeter - terminal black lead connected to led -) and open circuit in reverse bias (swap the probes around) when tested with a multimeter in diode check mode. The leds will not light from the diode test due t the current draw if the components around them.

The exact forward voltage drop will depend upon the characteristics of the red led but anything significantly different here indicates a problem.

0 voltage drop in both directions indicates a short across the led or the components it connected to. check for solder bridges and failed short transistors.

0L indicates in both directions means the led has dry solder joints or is blown.

 

using a multimeter in resistance mode check the leds

D23 close to the -360VDC psu input. In forward bias expect around 4M ohms. With the probes reversed you should get open circuit.

D24 is the led closest to the octal socket for the EL34 and should read about 1.8M in forward bias and about 1.9M in reverse. 

D1 is between in the 9 pin sockets but close to the heatsinks. This should read about 1.5M in both directions.

Readings in both directions of a few hundred K ohms or lower indicate the led may have been leaky or a transistor it is connected to has gone short circuit or a solder bridge.

 

Input terminal checks

With the multimeter in resistance mode and the - terminal connected to the amp ground, and the positive terminal connected to the + input of the amp you should get about 340K and the same for the - input. Naturally expect the same results for the other channel. If significantly different suspect incorrect resistors for R94, R7 (positive input). Or R95, R8 on the - input.

 

Output terminal checks

With the multimeter in resistance mode and the - terminal connected to the amp ground, and the positive terminal connected to the + output of the amp you should get about 0.5M and the same for the - output. Naturally expect the same results for the other channel. 

 

Valve Socket Checks.

With the multimeter in resistance mode and the - terminal connected to the amp ground, and the positive terminal expect the following resistances for both EL34s: (socket viewed from the top of the amp). (note the 0.5M reading may start lower and quickly ram up and the small capacitors in the circuit are charged by the multimeter).

 

For the 6922s (socket viewed from the top of the amp)

 

 

Diode Checks of transistors

Note it is possible for a transistor to fail in such a way it has very little gain but still has a diode drop and so diode checking transistors is not a foolproof measure of a transistors health. But 0V drop when not expected indicates a short etc.

Diode checks KSA1156 checks Q1, Q2, Q3

two of the ksa1156s (Q2 and A3) measure the same in diode mode the centre one (Q1)  differs slightly. All combinations of probes and pins on a transistor result in open circuit unless otherwise shown otherwise:

+ indicates positive multimeter probe attached,

- indicates negative multimeter probe attached

n/c indicates no probe attachment.

Note: for the other channel EACH KSA1156 is rotated 180 degrees about the y (vertical) axis.

Peak DCA in circuit testing Q1-3

The peak DCA 75 identifies all 3 transistors as PNP silicon. The centre with a hfe or around 28 and the other two with slightly higher hfe of 30 to 31. This test is reliable, anything significantly different indicates a problem. Note it is easier to do these tests from the underside although you can hook up the dca proe hooks to something like a sensepeek weighted self standing spring loaded probe tip assembly and test easily from the top (https://sensepeek.com/pcbite-20

 

Diode check and Peak DCA 75 for Q23P FJPF2145

This is reliably identified as a NPN with hfe 26 and is part of the virtual battery connected to the + output side of the channel. NOTE: It has a mirror image pinout compared to the ksa1156s.

Diode check and peak DCA 75 check for Q5 2SK216

This is reliably identified as a N channel mosfet with body diode and has a transconductance of 23.3mA/V. Although not part of the virtual battery it is directly connected to it. WANRING This transistor has a live mounting tab and must be fully insulated from the heatsink. A resistance check from the metal tab to the heatsink it is mounted on absolutely must read open circuit.

Diode measurement for this transistor does not show a stable reading. In this case dec indicates a reading that decreases over time.

 

The next two transistors in the heatsink row are Q23N and Q4 which perform exactly the same role and are exactly the same type of transistors as Q23P and Q5 respectively and should measure about the same I got 28mA/V for the Q4 so expect a little variance here).

NOTE: on the other channel the order of the 2SK216 and FJPF2145 are reversed.

 

Diode check and peak DCA 75 check for Q35 and Q36 FJPF2145

These are the current providers (driven by Q34) for a 10mA constant current source that feeds the input 6922s. The job is equally divided between them and they should measure the same. The DCA75 reliably identifies them as NPN silicon hfe 25-26

 

 

Diode check and peak DCA 75 check for Q15 and Q10 (10M90S)

These are identified out of circuit by the dca75 as N channel depletion mosfets. However, in this circuit they provide anode current and are wired in such a way the DCA75 can not identify them and reports no component. The DCA75 should not report any shorts and the metal mounting tab is live and absolutely must be insulated from the heatsink they are mounted on. Q15 and Q10 are identically configured and should measure the same. The gate and cathodes are connected together with resistors totalling only 400ohms so the diode drop voltages will be lower than for the other transistors also unlike the other transistors all combination of probes will result in a voltage drop. The voltage drop should be the same in the forward and reverse directions.

 

 

Diode check and peak DCA 75 check for Q28 and Q29 (KSA1156)

These are identified out of circuit by the dca75 as PNP silicon. However in this circuit they are wired in such a way the DCA75 can not identify them and reports no component. Each should measure the same.

 

 

Diode check and peak DCA 75 check for Q26 and Q27 (2SK216)

These transistors do not produce stable diode voltage drops.

 

Diode check and peak DCA 75 check for Q25 and Q24 (2SJ79)

These transistors are not easy to get a stable reading on. 

 

 

Diode and peak dca75 Check Balance servo Q37 - Q40 (MPSA06)

4 of the MPSA06es have their collectors and bases shorted together and so can be considered to only have two pins as far as diode checking concerned. 

NOTE for any transistor with shorted pins only two dca75 probes were used (one to the shorted pins and the other to the non shorted pin) to avoid the dca just reporting probes shorted. Understandably ni this case the dca75 can not identify the component as a transistor and instead reports a 9.87V zener. All other transistors are correctly identified and hfe provided.

 

 

Diode and peak dca75 Check Virtual Battery Q16 - Q18 (STN9360) and Q20 (MPSA06)

the other virtual battery in the channel (just to the left) should measure similarly and is identically laid out. Q20 pt 1 has collector and base shorted. All 4 stn9360 should diode measure similarly.

NOTE for any transistor with shorted pins only two dca75 probes were used (one to the shorted pins and the other to the non shorted pin) to avoid the dca just reporting probes shorted. Understandably ni this case the dca75 can not identify the component as a transistor and instead reports a 10V zener. All other transistors are correctly identified and hfe provided. The stn9360 are pnp and the remaining non shorted mpsa06 as npn

 

 

Diode and peak dca75 check Q30 and Q31 (FJPF2145)

Each transistor provides a separate 5mA current source and are directly connected to the virtual battery. With the led D24 across their base and emitter. Each is identically configured and should read the same. Dca75 reliably identifies both transistors as npn both hfe 18

 

Diode and peak dca75 check Q32 (on seperate small heatsink) and Q33 (FJPF2145)

this forms a 20mA current source which is controlled by Rv5 and sets the DC offset. DCA75 reliably identifies both transistors as npn both with hfe 26

 

Diode and peak dca75 check Q34 (FJPF2145)

Q34 provides control for Q35 and Q36 which creates a 10mA current source.

The dca75 says no component detected. 

 

<<< IN PROGRESS >>>

#include <usual_disclaimer.h>

#include <usual_high voltage_warnings.h>

 

 

Edited August 14, 2023 by jamesmking

[audiostar]

2 hours ago, jamesmking said:

#include <usual_disclaimer.h>
#include <usual_high voltage_warnings.h>

gcc -I. -c all_above.c -o working_amp.exec

[jamesmking]

Since the T2 needs a complex power supply I am also including a T2 power supply testing guide.

There are multiple options for the T2 power supply but the one I built is:

 

The gerbers can be found here:

<insert link>

 

The guide is intended for both pre-power on verification of an amp build, verification of voltages on power on and general troubleshooting. All tests are performed using a Brymen BM869s and peak DCA75. Using a different multimeter may effect the results slightly but you should still get similar behaviours and ball-park figures.

 

All tests are performed with the psu not connected to an amp and no mains transformer connected.

 

unfortunately many of the transistors are packed tightly together so it is not easy or really possible to get test probes onto all the legs of all the transistors from the top and so the verification guide will be separated into two halves: testing from the top and testing from the bottom.

Since all the high voltage power supplies are basically identical except for differing values for some resistors ad the pre regulator zeners I will cover one psu rai and the diode tests and dca75 tests should be the same for the other high voltage rails. (The 500V and above rails also have two caps in series with bleed resistors in parallel for the input smoothing and output smoothing to reduce the financial cost of the caps and increase the cap options) The low voltage rails are trivial - diode bridge, bulk cap, monolithic off the shelf regulator and smoothing cap and so will not be covered.

NOTE the 3w resistor which direct connects to the 10 volt reference will get hot (depending upon the supply rail) it can reach 80C+, If you look at my schematics above I suggest increasing the value of the resistor to reduce the current and decrease the temperature. Also raise it from the pcb as much as possible to keep it away from the nearby 0.1uF capacitor. 

NOTE if you use 0.1% precision resistors for the output voltage set resistors then they must be rated at above 250V for the 560V rail or (in my experience if you use 250V rated working voltage resistors their resistance will drift higher and higher with time and the 560V rail will eventually hit the zener string voltage of 600V and this will cause an uncorrectable dc offset in your T2 amp). 

 

Transistor and diode location Top

 

The two ksc5026 (T3 and T2) form a long tail pair with T3 input being the 10V reference and T2 input being the the voltage across R5 (and half of P1) in the voltage set string. To improve the performance of the differential amp T6 and T4 form a current mirror. R12 is the common resistor for the long tail. The output of the long tail comes from the collector of T3 and goes to T7.

 

Probe Tests from above

Diode test of the zener pre regulator D5 and D4

measuring across each separately, you will get about 0.6V drop, probes reversed the multimeter will register a steadily increasing voltage until it shows open. This is true of each zener in the string.

Diode test D2

you will get about 0.6v drop in one direction and an instant open in the other

Diode test D1

you will get about 0.52v drop in one direction and an instant open in the other

Diode test D3

you will get about 0.54v drop in one direction and an instant open in the other

 

Note it is possible for a transistor to fail in such a way it has very little gain but still has a diode drop and so diode checking transistors is not a foolproof measure of a transistors health. But 0V drop when not expected indicates a short etc.

Diode and dca75 tests of the 10m90s 

NOTE the 10m90s has a live tab and absolutely must be insulated from the heatsink it is mounted to.

Like the 10m90s in the t2 amp boards the 10m90s measures the same when the polarity of the probes has been reversed. Just like the T2 amp the peak dca75 can't identify this component in circuit and variously shows it as a low voltage zener ~ 1.6V or an led.

 

Diode and dca75 tests of the fqp8n80s

The peak dca75 can't identify this component in circuit and variously misidentifies it as two diodes or a diode and led.

 

 

Diode checks from underside

Transistor and diode location Bottom

 

 

Diode test T1 2N3904 NPN transistor.

forms the active part of the current limit circuit. It monitors the current through R15 and when the voltage drop across the resistor gets too high T1 starts to cut off the pass mosfet FQPF8M80C. If you get correct voltage output with no load but any load massively decreases the output down to about 75V suspect T1 has gone short circuit.

 

Diode and DCA 75 test T2 and T3 KSC5026 NPN transistor.

These form the long tail pair differential amp which compares the voltage reference against a portion of the output voltage controlled by the trimmer P1 and the series resistor  ladder R3, R4 R5. There should be very close  to 10V across R5 if not one possibility is R3 and R4 have too higher a resistance value, possibility from using resistors with a too low working voltage - especially in the -560V rail which puts more than 250V across each of R3 and R4 in the series ladder.

NOTE T2 is connected to a both ends of a cap and so this cap will slowly charge resulting in the diode test reading showing an increasing voltage drop until the multimeter finally displays open when the cap has charged to the same voltage as the meter outputs and so no current flows at all fooling the meter into thinking there is an open circuit. This does not occur with T3

The dca75 reliably and correctly identifies T2 as a NPN silicon transistor hfe 19 and T3 as a NPN silicon transistor hfe 2.

 

 

Diode and DCA75 test T4, T6 T7. Ksa156 PNP transistor.

T4 and T6 form  the current mirror for the long pair differential amp.

NOTE T4 has base and collector shorted together and so measures similarly to a single diode.

The DCA correctly and reliably identifies T7 as a PNP transistor hfe 135

The DCA obviously identifies T4 as a diode junction if you don't connect the 3rd lead, otherwise it just reports a short between two of the leads - which is correct.

The DCA correctly and reliably identifies T6 as a PNP transistor hfe 2

 

 

 

 

 

 

<<<work in progress>>>

 

 

 

Edited August 20, 2023 by jamesmking

[demonkuro]

I had a strange situation with my T2, when I set the R42 (22K) in all the batteries to 6.55V through the RV2 (10K), everything was fine, DC offset to ground was around -10V. When I measure the negative voltage of the battery, DC offset to ground suddenly becomes -150, I go to check R42, and I find that the R42 in the existing battery becomes 5.8V, some become 6V, and some become other values. I didn't adjust anything, and after constantly measuring the R42 of different batteries, they changed back to 6.55V, and DC offset to ground also returned to around -10V. Sometimes when I plug in the headphones for a while, DC offset to ground also gets bigger, and when I go to check the R42, I find that they also change, and after measuring different R42s, sometimes they change back to 6.55V, sometimes they don't. Sometimes DC offset to ground gets bigger when the power is turned back on, what's going on?

[Craig Sawyers]

Instability? Something in the circuit oscillating at beyond audio frequency?

[demonkuro]

13 hours ago, Craig Sawyers said:

Instability? Something in the circuit oscillating at beyond audio frequency?

I don't have an oscilloscope, I don't know if it has oscillations. If it does have oscillations, what should I do about it?

[sorenb]

22 hours ago, demonkuro said:

When I measure the negative voltage of the battery, DC offset to ground suddenly becomes -150, I go to check R42, and I find that the R42 in the existing battery becomes 5.8V, some become 6V, and some become other values

maybe start by checkin PSU rails at the AMP boards entry points, and check that all diodes are properly lit

[simmconn]

Use your most sensitive organs. Do not connect the headphones. The sound emanating from the tubes when T2 oscillates a good hearing test, in terms of both sensitivity and frequency response 😆.

Seriously, you could use the AC function of your multimeter. Most of them are good to tens of kHz, some good to a few hundred kHz. If something’s going on, there would be a good chance to see it on the AC meter once the reading settles.

The active battery circuit can be marginally stable. The capacitance of the test lead can tip the balance and cause it to oscillate. But R42 should usually be safe to probe on. If you see a lower voltage, you might want to check if the output voltage of the active battery (+/-) has dropped proportionally, i.e it still maintains its gain. From there you’ll know whether the reference chain or the amplifier part of the active battery should be checked next.

[demonkuro]

4 hours ago, sorenb said:

maybe start by checkin PSU rails at the AMP boards entry points, and check that all diodes are properly lit

All output voltages of the PSU are normal and all LEDs are on

[demonkuro]

4 hours ago, simmconn said:

Use your most sensitive organs. Do not connect the headphones. The sound emanating from the tubes when T2 oscillates a good hearing test, in terms of both sensitivity and frequency response 😆.

Seriously, you could use the AC function of your multimeter. Most of them are good to tens of kHz, some good to a few hundred kHz. If something’s going on, there would be a good chance to see it on the AC meter once the reading settles.

The active battery circuit can be marginally stable. The capacitance of the test lead can tip the balance and cause it to oscillate. But R42 should usually be safe to probe on. If you see a lower voltage, you might want to check if the output voltage of the active battery (+/-) has dropped proportionally, i.e it still maintains its gain. From there you’ll know whether the reference chain or the amplifier part of the active battery should be checked next.

When I approach the EL34 with my ear, there will be noise in one of them, and when I plug in the headphones, there will be noise in the headphones

When the outputs are measured using the multimeter AC function, their readings are constantly changing, like flashing.

When I use a multimeter to measure the voltage of R42, there is a screeching noise as soon as the black measuring pen (negative pole) touches R42, and they seem to be from tubes. The red pen (positive electrode) does not make noise when touched.

When the voltage reading of R42 changes, R42+ and R42- will not change in the same proportion, they will become different, one large and small, such as a 5.8v and a 6.4v, then the DC offset will become very large up to -80V or even -110V

[simmconn]

Please try to describe the procedure with more details.

1. Did you use a bench multimeter, or a portable one? The bench meter has slightly different characteristics between pos and neg leads, the portable one barely has any.

2. Did you put the pos lead on the pos node of R42 or the other way around? Did you connect the pos lead first or neg lead first?

3. when you mention “R42+ and R42-“, did you mean the R42s in the two active batteries of the same channel?

Chances are the reference chain on the active battery is marginal, such that the input impedance of your meter would throw it off. Check your meter, also check the K246 CCS. The voltage across the LED string and the voltage across RV2 shouldn’t change.

If the active battery oscillates, the DC voltage across R42s in the +/- arms of the same channel could be different. Do you have more than one meter? You can monitor the output of the active battery (across the test points) when you measure the voltage on R42.

 

[Kerry]

Depending on the Hfe of Q23, you might need to lower R39.  This could cause instability / oscillations in that area.

If you can hear the amp without the headphones in, then it's oscillating.  I agree that most multimeters should be able to pick this up (though maybe not accurately).

If the amp is wildly oscillating and has very high voltage swings it could possibly damage the amp.

[demonkuro]

6 hours ago, simmconn said:

Please try to describe the procedure with more details.

1. Did you use a bench multimeter, or a portable one? The bench meter has slightly different characteristics between pos and neg leads, the portable one barely has any.

2. Did you put the pos lead on the pos node of R42 or the other way around? Did you connect the pos lead first or neg lead first?

3. when you mention “R42+ and R42-“, did you mean the R42s in the two active batteries of the same channel?

Chances are the reference chain on the active battery is marginal, such that the input impedance of your meter would throw it off. Check your meter, also check the K246 CCS. The voltage across the LED string and the voltage across RV2 shouldn’t change.

If the active battery oscillates, the DC voltage across R42s in the +/- arms of the same channel could be different. Do you have more than one meter? You can monitor the output of the active battery (across the test points) when you measure the voltage on R42.

 

1. I use a portable multimeter.

2. (1) Only put neg lead, it has a screeching sound.

    (2) Put the neg lead first, which has a screeching sound, and then put the pos lead on the other end of the R42 and the screeching sound disappears.

    (3) Put the pos lead first, there is no screeching sound, and then put the neg lead on the other end of the R42, and there will be no screeching sound.

3.Only when measuring R42 in the R channel, near the E88CC (6922) tube, a screeching sound is generated.

When the machine is cold, everything is normal when it is turned on, and all output DC offset is only about -7V.

After a while, when the machine gets hot, the DC offset output of the R channel will become larger and more than -100V. The voltage of the two batteries of the R channel will decrease and become different, the voltage of the two R42 will also decrease, and the voltage of the two R42 will become different. One R42 becomes 6.2V, and the one close to E88CC becomes 5.8V, they are irregular and do not drop to this value every time.

Even when the machine is cold, as long as you measure the R42 close to the E88CC, the voltage of the battery will immediately decrease, the other R42 will also reduce the voltage, and the DC offset of the output will become -100V.

4.I used the original T2 PCB, but based on JoaMat's T2 modified I replaced some parts. I replaced 2sa1486 (Q16, Q17, Q18, Q19) with stn9360. Replaced all 2SC3381 with 2 MPSW06. Replace 2sa1486 (Q1, Q2, Q3) with KASA1156.

Q6, Q7, Q11, Q12 have not changed, they are still 2sa1486.

Are these changes in mine that make it oscillate?

I only have one multimeter and I can't measure multiple locations.

Thank you，simmconn.

5 hours ago, Kerry said:

Depending on the Hfe of Q23, you might need to lower R39.  This could cause instability / oscillations in that area.

If you can hear the amp without the headphones in, then it's oscillating.  I agree that most multimeters should be able to pick this up (though maybe not accurately).

If the amp is wildly oscillating and has very high voltage swings it could possibly damage the amp.

Thank you, Kerry, for reminding me. My Q23 HFE is 48~50, should I choose a lower HFE? Like 30? I'll try it.

[simmconn]

12 hours ago, demonkuro said:

1. I use a portable multimeter.

2. (1) Only put neg lead, it has a screeching sound.

    (2) Put the neg lead first, which has a screeching sound, and then put the pos lead on the other end of the R42 and the screeching sound disappears.

    (3) Put the pos lead first, there is no screeching sound, and then put the neg lead on the other end of the R42, and there will be no screeching sound.

3.Only when measuring R42 in the R channel, near the E88CC (6922) tube, a screeching sound is generated.

That sounds normal. Probing the R42 (-) injects noise to the diff amp. Connecting R42 (+) to the same meter reduces the injected noise.

Also the active batteries will not be *exactly* the same in term of stability even if you match the parts.

13 hours ago, demonkuro said:

After a while, when the machine gets hot, the DC offset output of the R channel will become larger and more than -100V. The voltage of the two batteries of the R channel will decrease and become different, the voltage of the two R42 will also decrease, and the voltage of the two R42 will become different. One R42 becomes 6.2V, and the one close to E88CC becomes 5.8V, they are irregular and do not drop to this value every time.

To what level did the voltage of the two batteries of the R channel decrease? What are the voltages at the (+) and (-) of the two batteries when that happens?

Remember I suggested you to check 1) the K246 CCS 2) whether the active battery still maintains the gain when the output decreases [V(+)-V(-)]/V(R42).

12 hours ago, demonkuro said:

Even when the machine is cold, as long as you measure the R42 close to the E88CC, the voltage of the battery will immediately decrease, the other R42 will also reduce the voltage, and the DC offset of the output will become -100V.

If you only have one multimeter, how did you notice the voltage of the battery change when you probe R42? Both the input and output stage offset servo and the global NFB affects the DC offset. So it's important to see what causes the offset to be off beyond their correction capabilities. To be safe, also check the input impedance of your meter.

13 hours ago, demonkuro said:

4.I used the original T2 PCB, but based on JoaMat's T2 modified I replaced some parts. I replaced 2sa1486 (Q16, Q17, Q18, Q19) with stn9360. Replaced all 2SC3381 with 2 MPSW06. Replace 2sa1486 (Q1, Q2, Q3) with KASA1156.

Q6, Q7, Q11, Q12 have not changed, they are still 2sa1486.

Are these changes in mine that make it oscillate?

I only have one multimeter and I can't measure multiple locations.

Thank you，simmconn.

Thank you, Kerry, for reminding me. My Q23 HFE is 48~50, should I choose a lower HFE? Like 30? I'll try it.

The STN9360 generally have higher HFE than the 2SA1486. Everything else being equal, the stability will be a little worse. Using Q23 with lower HFE may help with stability, but I feel that the oscillation and the thermal-related stability may be two separate issues here.

 

I don't quite understand that some people would spend thousand of $ on parts, building a unit that has a fair market value of close to 10k, while not willing to invest a few hundred $ on quality tools and equipment. If we know Stax only uses a multimeter to build and test T2, we'll probably avoid their amps like plague. How did that become ok when some lucky individuals do the same?

[jamesmking]

54 minutes ago, simmconn said:

I don't quite understand that some people would spend thousand of $ on parts, building a unit that has a fair market value of close to 10k, while not willing to invest a few hundred $ on quality tools and equipment. If we know Stax only uses a multimeter to build and test T2, we'll probably avoid their amps like plague. How did that become ok when some lucky individuals do the same?

I hope he is not using this multimeter. It has a built in transistor test feature so you know its quality....

resistance measurement only up to 2Mohm, and ac bandwidth of an insanely narrow 45hz to 450hz. 500V max input and is not even true rms... all for £2.95 before tax.

https://cpc.farnell.com/duratool/d03046/multimeter-digital/dp/IN07220?mckv=s_dc|pcrid|426684131405|kword||match||plid||slid||product|IN07220|pgrid|100371162238|ptaid|pla-1751967123517|&CMP=KNC-GUK-CPC-SHOPPING-9262013734-100371162238-IN07220&s_kwcid=AL!5616!3!426684131405!!!network}!1751967123517!&gclid=CjwKCAjwu4WoBhBkEiwAojNdXqPLfzyh0bPlKygDGgTcWuNmZFHencRS08mstzeQNiGSrgSmFODsSBoCkpYQAvD_BwE

😬 💥 😱 🚑 

 

 

Edited September 13, 2023 by jamesmking

[Kerry]

You should use a higher Hfe. What you have is ok, but if lower R39 to around 40k to 50k. Highest value that is stable. 

[simmconn]

May I ask the reason for reducing R39? Is it because of the STN9360 HFE drop under very low collector current? I did find that STN9360 and 2SA1486 response differently to the compensations I added to improve stability, but that’s probably more due to their different Cob.

Thanks!

[jokerman777]

So I'm trying to adjust batteries on one channel for now, and was trying to follow Kerry's method. 

I can set the voltage across the two 22k resistors to be around 6.5V by turning the 10k pot fine, but then it seems turning the 2k pot only affect the positive (~200V) battery voltages and the negative voltage sits still at -548V. I tried turning away the 10k pot a little bit too and neither does it seem to affect the negative voltage but it does change the positive as well, so for now I can have +200V/-548V for this channel. 

All LEDs in this channel currently light up fine, but weirdly sticking probes across the 22k resistor can dim the some LEDs and cause some squeak noise on the tubes..

Is this behavior expected at all that I shall just move on and adjust the rest or does it look like something is off?

 

[JoaMat]

Only positive voltage is expected to change.

Set voltage cross 22K resistor to 6.5V and don’t change that unless a good reason. Battery voltage of 740V is mostly a good value. But if 748V works be happy.

[JoaMat]

New device package in Proteus.

Generic output tube footprint for DIY T2. Accepts EL34, EML 20B-V4 and 300B. Will see if fab house can fix this.

[kevin gilmore]

clever idea but you then need a jumper block.

[JoaMat]

The idea is to populate the five 0805 pads according to the table.

disclaimer: if things blow up don't blame me.

Edit: Intention is to have a board compatible with the three tube types. When building the amplifier you decide for which tube. So, no "tube type rolling" here.

Edit: Table changed.

 

Edited October 8, 2023 by JoaMat

[simmconn]

No grid resistor R53 for the 300B?
I wonder why the original design uses such a high value 22k. It’s going to interact with the Cga and cause some interesting negative feedback at high frequencies.

Also, with the 300B, the max output amplitude could be quite limited. You’ll need about -100V bias to get the 300B to 400V @ 10mA on the plate. But the max negative grid swing is limited to about -160V due to the 2Sk216 Vds max. The result is limited positive output swing, or operating the 2Sk216 at risk.