DIY mini T2 Build Thread

[Laowei]

Nice to see they got married. That’s legal now in most countries around the world.

[gepardcv]

That is so impressively miniaturized that pairing it with the beastly T2 PSU and its three huge transformers seems just wrong.

[Pars]

Yeah, probably. Unfortunately, unless you want to go with SMPS (or even can), the PSU is where the physics rubber meets the road, so to speak...

[mwl168]

See the zip file below; JoaMat kindly put together a list of component value changes for the mini T2 that the builders can apply to the V1 boards received from me. These values will put your mini T2  close to the operation points of the original DIY T2. 

Following the values on the silkscreen without changes will still result in an working amp. 

The zip file contains a complete list of the changes and pictures of the boards outlining the changes. I also attached a screen shot that shows the list of changes.

Thanks very much JoaMat for doing this!

Changes mini T2.zip

Edited March 1, 2019 by mwl168

[chinsettawong]

From the schematic, the mini T2 will be using +220V, +400V, -460V, +15V and -15V.  I'm thinking of using the PS board from my prior KGSSHV build.  Is it possible to use a simple voltage divider to get the +220V from +400V?  How much current does each voltage draw?

 

 

[gepardcv]

I wonder if there's a good way to adapt the GRHV. Crank the negative rail to -460V, keep the positive rail at +400V, and (if possible) use a voltage divider to make a +220V. Way simpler and smaller than the T2 PSU.

[jose]

17 hours ago, chinsettawong said:

From the schematic, the mini T2 will be using +220V, +400V, -460V, +15V and -15V.  I'm thinking of using the PS board from my prior KGSSHV build.  Is it possible to use a simple voltage divider to get the +220V from +400V?  How much current does each voltage draw?

 

 

I´m not sure but I think that it´s not possible. Used a specific PSU is a better idea. 

[MLA]

+1; JoeMat suggested a separate PSU for the +220V in a previous post somewhere.

I'm planning on using the PSU Kevin created for the tube input board for the +220V (have to check it works first though :)).

[jose]

I have to look at this but maybe with + rail of the GRVH (Split pcb´s) maybe I can get + 220Vdc

[mwl168]

18 hours ago, gepardcv said:

I wonder if there's a good way to adapt the GRHV. Crank the negative rail to -460V, keep the positive rail at +400V, and (if possible) use a voltage divider to make a +220V. Way simpler and smaller than the T2 PSU.

I've discussed this topic with JoaMat early on. His advise is that the +220VDC be supplied by its own regulator instead of using a drop resistor from the B+. The top device, HN4A51J, is a dual transistor chip and need to be careful not to overheat it.

The draw on the +220VDC rail should be around 20-30mA both channels combined.

 

[JoaMat]

With R1 and R3 at 270R you get 4 mA through each current source plus 0.5 mA LED current. Put a resistor of 20K/2W between +400V and +220V pad on the amplifier board - gives approximately 224V. Someone try this and give a report? I’m on a Ski holiday.

[JoaMat]

Maybe it’s not wise to entirely rely on a single 20K resistor. A better way probably is to use Zener diodes to achieve 220V as Kevin does with the KGSSHV-Tube-Sandwich.

Another way could be using mosfet like Kerry did in below LTspice simulation – see U10, R16 and R22.

Ref. Kerry post https://www.head-case.org/forums/topic/12949-stax-t8000-clone-well-sorta/?do=findComment&comment=764504

Edited March 7, 2019 by JoaMat

[JoaMat]

Back from skiing. Zoom in on the Vishay CPF3 resistor and you can read 20K. 400 volts in at the blue wire. The meter is in Min Max mode and it went up to 225.5 V at start up and then immediately down to what you see in Picture.

Seems to work. I’ll change the other channel and will test this for some time.

I have run the amplifier on a unchanged T2 PSU with +/-500 V and +250 V for a while and that also seems to work alright.

 

Edit: Both channels now with 20K resistors to achieve the +220V. 220.3V respective 219.6V with PSU +400V trimmed to 401.9V on the clocks.

Edited March 10, 2019 by JoaMat

[JoaMat]

I haven’t run the amplifier with 20K resistors for more than a day. But after switching power on and off frequently I do think this will work. With R1 and R3 at 270 ohm I get 9.0 to 9.1 mA at +220V (each board) and with PSU positive rail set to 400 V a 20K/3W resistor seems to fit perfect.

 

On ‎3‎/‎5‎/‎2019 at 10:22 PM, gepardcv said:

I wonder if there's a good way to adapt the GRHV. Crank the negative rail to -460V, keep the positive rail at +400V, and (if possible) use a voltage divider to make a +220V. Way simpler and smaller than the T2 PSU.

I think gepardcv idea will work and in my opinion it’s absolutely worth trying. Current demand positive rail is approximately 80 mA and negative is 60mA.

My earlier statement that a +220 V supply is preferred might be a bit exaggerated.

[chinsettawong]

@JoaMat Do you recommend that we use the original T2 voltages or +400V/-460?

 

[JoaMat]

Well, I rather not make recommendations but I would go for +400V/-460V and add one 20K/3W resistor on each board between +220V and +400V pads.

I think I want something like this…

...for power supply.

Edited March 11, 2019 by JoaMat

[Kerry]

I've tested this supply and it works nicely 

[JoaMat]

It looks like the LT1021 is abandoned by most supplier. Mouser is out of stock but claims that it’s on order. There is another 10 volt reference LT1236-10 that I think should be a good alternative. I’ve tested it in LTspice model by Kerry (jump here for reference and more info) and it seems to do the same as LT1021-10 but with 0.2 mA lower current with same current set resistor. Someone has to verify LT1236 works.

I guess Kerry needs a couple of LTxxxx in above PSU.

[G600]

We definitely need a group buy for those PS.

All-in-one PSU for the miniT2 kept mini (apart from the transformers, but that's another story).

[gepardcv]

I've bought LT1021 from Digi-Key in the past. Doesn't look like there's a ton in stock, but I think it comes and goes.

[Inu]

How about LT1021-7 with 510 - 560 ohm? 

[Pars]

I've always bought mine directly from Linear...

[JoaMat]

LT1021-7 is worth trying – in LTspice it acts like LT1021-10 and LT1236-10.

The LED TLMS1100 is a tiny thing and personally I think that’s the most tricky part to handle and solder. I started to use it in battery daughter board for T2 and I did buy a hundred pieces right away. It might be a good idea to buy some extra in case if you lose one (on floor, overheat when solder or what so ever). Now days I solder in the LED first and then check it with DMM in diode test mode, positive test pin on anode and negative on cathode and the LED will glow if everything is right (no guarantee for bad soldering though - by own experience). It will not glow with transistor and resistors in place.

23 hours ago, G600 said:

We definitely need a group buy for those PS.

All-in-one PSU for the miniT2 kept mini (apart from the transformers, but that's another story).

We send the designer a bottle of Scotch each and see what happens…

Edited March 13, 2019 by JoaMat

[Kerry]

For the GRHV78XXX I would recommend to have the full board assembled as part of the group buy.

Yeah, 0603 LEDs are fun

[mwl168]

52 minutes ago, Kerry said:

For the GRHV78XXX I would recommend to have the full board assembled as part of the group buy.

Yeah, 0603 LEDs are fun

Happy to organize a GB if there is enough interest. 

Please raise your hands and state quantity desired here if interested.