Craig Sawyers Posted December 31, 2010 Report Posted December 31, 2010 Hokay - I've decided to go with 750 ohms. I have some MRS25 in 750 ohms. These are truly tiny 600mW resistors, and will fit vertically between the hole in the board and the gate of the K216's with no difficulty. Measured one on my ancient (but super accurate) HP R-X bridge, and even at 150MHz it still meausures 710 ohms in parallel with 0.3pF - so pretty much an ideal resistor.
kevin gilmore Posted December 31, 2010 Report Posted December 31, 2010 No ferrite beads on the original. No evidence of oscillation in my unit, now about a year old. andy has a problem in one channel of static that comes and goes, he has a scope, so will see if there is any oscillation... A ferrite bead might not be a bad idea.
Craig Sawyers Posted December 31, 2010 Report Posted December 31, 2010 No ferrite beads on the original. No evidence of oscillation in my unit, now about a year old. andy has a problem in one channel of static that comes and goes, he has a scope, so will see if there is any oscillation... A ferrite bead might not be a bad idea. The problem is that Mosfet self oscillation is bewlideringly prone to both precise layout almost at the millimeter level, exactly where the bypassing is in relation to the transistor, and the gm of the transistors. There is a rather wide tolerance in gm, so perhaps an unlucky batch with high gm might be a trigger. There is an excellent Renesas apps note (simply called "Power Mos Fet Application Note" - 57 pages) which covers more Mosfet death mechanisms than I ever knew existed, parasitic self oscillation being only one. Maybe there could be some interaction with the input tubes which might explain the noisyness with some tubes that both you and Inu have mentioned. I don't have any stock of beads, so I'll go with the resistor and keep the value the same as for the J79.
Craig Sawyers Posted December 31, 2010 Report Posted December 31, 2010 Right - several hours of T2 rebuild with measured semiconductors, and the beast is ready to roll again. However, I am now going to drink beer and cook. And we're out tomorrow visiting friends. So it will be Sunday when I do final checks, plug in the mains and hopefully not do this to myself.
Craig Sawyers Posted January 2, 2011 Report Posted January 2, 2011 Latest hold up - LED's. D6,7,10,11 popped when I was in the crap C3675 phase. Replaced them with stock 5mm red LED's. But before switching on I thought I'd check the voltage. Ones specified by KG measure 1.8V at 0.5mA - which then gives the (calculated) currents on the schematic exactly. The ones I replaced them with measure 1.6V at 0.5mA. Does not seems much, but it shifts the operating point of the EL34's by 20%. So I am procuring more LED's, which puts a few days delay into the turn on. Ought to have bought a shed load more in the first place, since they are cheap. This sucker is going to be 100% according to design before it sees volts again.
Craig Sawyers Posted January 3, 2011 Report Posted January 3, 2011 I've really been pasting this list thick and fast recently. But for anyone interested, here are a couple of spice models. The first is for the 2SA1413, which is a different package for the 2SA1486, and is a Sanyo model. The second is for the 2SC3675, generated by a downloaded SPICE parameter calculation spreadsheet downloaded from excel , with input data measured by me. That is essentially a dc model only, with no description of electrode capacitances, and nothing to do with inverse active mode. .MODEL 2SA1413 PNP (IS=153.22E-15 + BF=73.199 + VAF=104.34 + IKF=1.3862 + ISE=153.22E-15 + NE=1.3075 + BR=.357 + VAR=100 + IKR=3.7415 + ISC=63.966E-12 + NC=1.7723 + NK=.92315 + RB=.73481 + CJE=1.3316E-9 + VJE=.77186 + MJE=.38206 + CJC=140.65E-12 + VJC=1.1825 + MJC=.54838 + TF=5.0775E-9 + XTF=2.5439 + VTF=16.727 + ITF=193.71 + TR=30.928E-9 + NR=1 + NF=1 + TRC1=0 + TRB1=0 + XTB=1.5) .model 2SC3675 NPN (VAF=6000 IS=1.00290528476071E-14 +NF=.934004260398326 BF=150 NE=1.11055219144042 +ISE=9.61916074296322E-15 IKF=.1 +TF=2.6525846254073E-08 )
Craig Sawyers Posted January 3, 2011 Report Posted January 3, 2011 Equipped with my shiney new Spice models, I've had another go at the battery, this time with the bottom current source around Q30 modelled too. I've been investigating what RV1 and RV2 do in preparation for turning the thing back on in the coming days. RV1 does nothing much over quite a lot of its range. The model I have of the 2SK246 seems to have an Idss of 10mA, so turning RV1 *anticlockwise* initially has the battery completely off and at maximum voltage (790V or so) for the first 6.5k of its travel. The battery voltage then falls at 34V per turn of the pot. In the model, 740V is reached at 6.9k. This would appear to be the "coarse" control. RV2 is essentially linear at 3.7V per turn (ie around ten times more sensitive than RV1), and would appear to be the "fine" control. It operates over a total range of 90V. Clockwise reduces the voltage. So the strategy seems to be to set RV1 and RV2 to mid travel. Adjust RV1 to get withing striking distance of 740V (+/- 25V or so) and then fine trim with RV2. At least that is what the modelling says. I'll try it out in the coming days and report back.
Inu Posted January 5, 2011 Report Posted January 5, 2011 Here's a question for Inu and Kevin. Is there any indication of the power Mosfets oscillating? Or bursts of oscillation at certain points in real signal? Reason I mention it, and I should have recalled this from power amp designs, they have a horrid tendency to burst into high frequency self oscillation. There was one generation of Trace Elliott bass guitar amp in which the power mosfets physically exploded through this effect (although at 1kW output there was LOTS of energy driving them). I spotted that on the K216's on my curve tracer, but the patch leads were around 2 inches long to the test fixture and certainly had an effect. In the T2 design, there is a 750 ohm resistor in series with the J79 gates (and KG has it tight up to the gate, which is exactly right), which is a classic way to stop this (the old grid stopper idea), and kind of suggests that it was introduced to stop self oscillation of the J79 in the orignal circuit layout. But there is no such protection on the K216's. So the question is - do we need to do anything? Either a resistor in series with the K216's gate or a ferrite bead on the gate lead? No oscillation but I added 500ohm to the K216 Gate. Picasa Web Albums - inuponken - T2
Craig Sawyers Posted January 5, 2011 Report Posted January 5, 2011 No oscillation but I added 500ohm to the K216 Gate. Ah - that is interesting, Inu. I've added 750 ohms to mine as a precaution; we were clearly thinking along the same lines. Incidentally, I put a spreadsheet of battery voltage against RV1 and RV2 position from a spice simulation here http://www.tech-enterprise.com/tekstuff/Batteryvoltage.xlsx
Kerry Posted January 5, 2011 Report Posted January 5, 2011 Which of the K216's are you doing this too? Just the ones feeding the output stage (Q26, Q27) or are you also doing it to the ones feed the active battery (Q4, Q5)?
Craig Sawyers Posted January 5, 2011 Report Posted January 5, 2011 Which of the K216's are you doing this too? Just the ones feeding the output stage (Q26, Q27) or are you also doing it to the ones feed the active battery (Q4, Q5)? I've done both. Both are active - all the battery does is apply a 740V DC level shift; from an ac point of view the battery is short circuited via C3.
Craig Sawyers Posted January 6, 2011 Report Posted January 6, 2011 WHOOOOOO! First channel operating!!! RV1 and RV2 do exactly what the simulation and spreadsheet says, as is my suggested way of adjustment. Outputs sitting at -4.5V and batteries at 743V. I can probably get the output offset lower than that, but I'll now concentrate on the other channel for a while. 1
Craig Sawyers Posted January 6, 2011 Report Posted January 6, 2011 WHOOOO! Squared! Second channel now operating. Adjusting those pots is pretty trivial having worked out what they actually do - and having halfway decent silicon in there. I'm going to award myself a run, some lunch, and then put the front panel on and have a listen 1
Fing Posted January 6, 2011 Report Posted January 6, 2011 Congrats Craig! It's been of great interest to read about the problems plaguing your build and what you did to troubleshoot and solve them - thanks for sharing about it.
Knuckledragger Posted January 6, 2011 Report Posted January 6, 2011 x2. This thread has been a fascinating read for me, and I'm an utter neophyte when it comes to the Dark Arts of Soldering.
Craig Sawyers Posted January 6, 2011 Report Posted January 6, 2011 Yeah Craig! What tubes are you using? At the moment a used Edicron quad that I had lying around that still measures pretty good. I've a good set of original Mullards that will probably end up in there, or I might get another set of Winged C (I've a set of those in the BH, and they sound very good indeed) and keep the Mullards for Sunday Best. I've got a set of Pearl tube coolers too - evening out variation in the envelope temperature and reducing it double tube life - something that was researched by Mullard etc back in the 50's. 6DJ8's are Mullard's pilfered from a wrecked Tektronix scope that measure right on the button.
Pars Posted January 6, 2011 Report Posted January 6, 2011 WTG Craig! And yes, this thread has been interesting to follow; top notch work.
deepak Posted January 6, 2011 Report Posted January 6, 2011 WHOOOO! Squared! Second channel now operating. Adjusting those pots is pretty trivial having worked out what they actually do - and having halfway decent silicon in there. I'm going to award myself a run, some lunch, and then put the front panel on and have a listen Very nice! You're well into dinner time, don't keep us hanging let us know how it sounds
Inu Posted January 6, 2011 Report Posted January 6, 2011 WHOOOO! Squared! Second channel now operating. Adjusting those pots is pretty trivial having worked out what they actually do - and having halfway decent silicon in there. I'm going to award myself a run, some lunch, and then put the front panel on and have a listen Congratulation!
Voltron Posted January 6, 2011 Report Posted January 6, 2011 Very nice! You're well into dinner time, don't keep us hanging let us know how it sounds Congrat Craig! And I second this^ emotion!
kevin gilmore Posted January 6, 2011 Report Posted January 6, 2011 If this is anything like when Inu got his finished, and decided to play hookey from work at lunchtime... We may not hear from craig for a few hours. And he will have a smile on his face that won't wash off.
deepak Posted January 6, 2011 Report Posted January 6, 2011 I wish no word “2SC3675” or “2SK216” in his next post.
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now