TT/Mega STE power supply headaches

Repairs & upgrades.
User avatar
exxos
Site Admin
Site Admin
Posts: 4543
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: TT/Mega STE power supply headaches

Post by exxos » Fri Apr 20, 2018 9:48 pm

If ZD3 was like 6Volts, then the 5V would have OVP anyway...

Though assuming that was true.. why bother taking a tap of the transformer as well anyway :shrug:

To tried to think right now.. but could be some forum of funky circuit where the negative rail swings positive turning on Q2 as the PSU turns off to discharge it.. I guess if you have a working PSU, measure the voltage on the bottom end of R28 , power up and check voltage, power down checking voltage.. see if the voltage jumps up when the PSU is turned off..
4MB STFM 1.44 FD- VELOCE+ 020 STE - 4MB STE 32MHz - STFM 16MHz - STM - MEGA ST - Falcon 030 CT60 - Atari 2600 - Atari 7800 - Gigafile - SD Floppy Emulator - PeST - HxC - CosmosEx - Ultrasatan - various clutter

https://www.exxoshost.co.uk/atari/ All my hardware guides - mods - games - STOS
https://www.exxoshost.co.uk/atari/last/storenew/ - All my hardware mods for sale - Please help support by making a purchase.

damanloox
Posts: 23
Joined: Mon Apr 16, 2018 7:10 pm

Re: TT/Mega STE power supply headaches

Post by damanloox » Sat Apr 21, 2018 9:53 am

I replaced SHR2 yesterday (with a new one) and still no joy - exactly the same behaviour...
I measured some more voltages and there's something I don't understand - I can see SHR2 has 1.8V on the cathode. So the Zener should prevent Q2 from turning on but the voltage on Q2 base is 0.7 and 1.1V on the collector... Which would suggest it os not open (at least not fully)... And if it's closed - why the behaviour is different than when Q2 is removed...?
I think I've had enough - ordered a few elements and will replace all transistors/diodes etc... I'll also replace C22 just in case...

User avatar
exxos
Site Admin
Site Admin
Posts: 4543
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: TT/Mega STE power supply headaches

Post by exxos » Sat Apr 21, 2018 10:17 am

0.7V is enough to turn the transistors on... 1.1V on C-E is the voltage drop across the transistor (pulled high by resistor to 5V). That transistor should not be turned on.. its why I suggest a second diode to drop the voltage 0.5V...
4MB STFM 1.44 FD- VELOCE+ 020 STE - 4MB STE 32MHz - STFM 16MHz - STM - MEGA ST - Falcon 030 CT60 - Atari 2600 - Atari 7800 - Gigafile - SD Floppy Emulator - PeST - HxC - CosmosEx - Ultrasatan - various clutter

https://www.exxoshost.co.uk/atari/ All my hardware guides - mods - games - STOS
https://www.exxoshost.co.uk/atari/last/storenew/ - All my hardware mods for sale - Please help support by making a purchase.

damanloox
Posts: 23
Joined: Mon Apr 16, 2018 7:10 pm

Re: TT/Mega STE power supply headaches

Post by damanloox » Sat Apr 21, 2018 4:47 pm

I replaced most of the components of PG circuit (SHR2, ZD3 and also PC111 just to be sure). I'm only waiting for transistors and will replace them as well.
Unfortunately no change in the behaviour... One thing that is really surprising is that there's +48V on cathode of D10 (ie. R25)... Any idea why it's there? And how?

czietz
Posts: 54
Joined: Sun Jan 14, 2018 1:02 pm

Re: TT/Mega STE power supply headaches

Post by czietz » Sat Apr 21, 2018 8:22 pm

I only just now noticed that you moved the discussion to this forum. You could have mentioned... given that I as trying to help you at AF. Since you simply vanished from that thread, I assumed you had fixed your problem.

Anyway, it is perfectly normal to have >40 V at the cathode of D10. (E.g. in my TT PSU the voltage is about 45 V.) This part of the circuit, including SHR2, Q2 and ZD3 handles the case of power fail -- see the technical specs of the PSU. Notice how D10 is oriented differently than all other diodes (D7..D9) on the secondary side? This is because it is supposed to be forward biased whenever the primary side of the transformer is being "charged". (The other diodes are reverse biased in that case; see the description of a flyback converter on Wikipedia if you're not familiar with the basics.)
As soon as the power fails, D10 doesn't provide any current anymore, the voltage over C22 rapidly decreases and in the end Q2 pulls the power good signal low.

damanloox
Posts: 23
Joined: Mon Apr 16, 2018 7:10 pm

Re: TT/Mega STE power supply headaches

Post by damanloox » Sat Apr 21, 2018 8:41 pm

Sorry - i forgot to follow AF thread...
I'm desperate at this stage - I replaced all components of PG circuit (except Q2, Q3 and Q4 as I'm waiting for the delivery and except resistors although I replaced R28 and R29. I also desoldered and measured r34 and it's exactly 452 ohms).
No change - all voltages (+/-12 +/-5)seem to be ok but PG only comes up after about 40 seconds.
I'm so desperate I'm actually considering removal of Q2 completely (as without it PG is ok within a second) but I'm afraid of those spikes at 5v and 12v at the startup.
I still can't understand why Q2 seems to be open (apparently - considering the fact that Vc is 1.1v and PG is 0...) and (apparently) closes after about 40 seconds..
And again - (once pg is up) if I turn TT off and power it back up within a few minutes - PG comes up pretty much instantly.. doesn't it look like a capacitor issue...?

User avatar
exxos
Site Admin
Site Admin
Posts: 4543
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: TT/Mega STE power supply headaches

Post by exxos » Sun Apr 22, 2018 10:41 am

As I said already.. Q2 is ON preventing c23 from charging. You need to do as I suggested to diagnose the problem. Randomly changing parts is not the answer!
4MB STFM 1.44 FD- VELOCE+ 020 STE - 4MB STE 32MHz - STFM 16MHz - STM - MEGA ST - Falcon 030 CT60 - Atari 2600 - Atari 7800 - Gigafile - SD Floppy Emulator - PeST - HxC - CosmosEx - Ultrasatan - various clutter

https://www.exxoshost.co.uk/atari/ All my hardware guides - mods - games - STOS
https://www.exxoshost.co.uk/atari/last/storenew/ - All my hardware mods for sale - Please help support by making a purchase.

czietz
Posts: 54
Joined: Sun Jan 14, 2018 1:02 pm

Re: TT/Mega STE power supply headaches

Post by czietz » Sun Apr 22, 2018 11:27 am

Well, at least with all that replacing you have really narrowed it down to Q2.

During the 40 seconds it takes for power good to become high, do you see any voltages (e.g. on C23) slowly change or does everything suddenly switch after 40 seconds? As for why the PSU works normally when switched off and immediately back on, there are two plausible explanations: 1. capacitors still charged, 2. thermal effects. (You could rule out the latter using a can of freeze spray.)

damanloox
Posts: 23
Joined: Mon Apr 16, 2018 7:10 pm

Re: TT/Mega STE power supply headaches

Post by damanloox » Sun Apr 22, 2018 2:39 pm

@exxos: I will try another diode (although I'mo not sure if 0.6v will change anything as there's 47v there and r26/r34 cut the voltage down significantly..

I took apart the other (working) TT/PSU and here's comparison of the measurements (working one is measured when PG is high as it only takes 0.5s for it to go high). Bad PSU has PG low all the time during measurements
- working PSU
D10+ 47.7
R25 46.2
R26/R27 5.43
SHR2 ref 2.7
SHR2- 1.88
Q2 b 0

bad
D10+ 47.5
R25 46.4
R26/R27 5.45
SHR2 re 2.66
SHR2- 1.83
Q2 b 0.6

Everything is pretty much the same except... Q2 base... on good one there's 0 (which I'd expect) but on bad one it's 0.6... Zener is a suspect but... I have changed ZD3 (and I tested one that I removed and it was OK). Replacement zener was brand new nexperia brand (from farnell). And yes - I made sure it's soldered in correct direction...
Is there any explanation of that 0.6v on Q2 base..? As I said - I'm waiting for new transistors but when I removed Q2 I did a quick test and it seemed to be OK (but I only tested it as two separate diodes as I don't have any specialised transistor testing kit)

@czietz: I will test and see it the voltage rises slowly or suddenly later this evening...

User avatar
exxos
Site Admin
Site Admin
Posts: 4543
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: TT/Mega STE power supply headaches

Post by exxos » Sun Apr 22, 2018 2:57 pm

You could try a 1k pot in place of the 543R resistor and just try a few ohms higher and lower to alter the bias voltages. Though if the 5v is to high that could effect things as well.

Using the diode in various places may also fix the problem, but we need to eliminate each possibility one by one until we find which line has a issue. But 1k pot be easier.

In my PSUs the brand of zener can make or break the protection, its sensitive to tollerences. As there is a 543R resistor in the circuit, that's a very specific value which makes me think it also has tollerences issues...
4MB STFM 1.44 FD- VELOCE+ 020 STE - 4MB STE 32MHz - STFM 16MHz - STM - MEGA ST - Falcon 030 CT60 - Atari 2600 - Atari 7800 - Gigafile - SD Floppy Emulator - PeST - HxC - CosmosEx - Ultrasatan - various clutter

https://www.exxoshost.co.uk/atari/ All my hardware guides - mods - games - STOS
https://www.exxoshost.co.uk/atari/last/storenew/ - All my hardware mods for sale - Please help support by making a purchase.

Post Reply