D&R4000 PSU Issues

[Siegfried Meier]

Ok, so testing around some more, I pulled the 723 that was in the +18v side, and I still get the same readings as before...

Umm...what?  How is this even remotely possible?  I've built up a few PSU's before, and if you pull the voltage regulator, you get no voltage...how am I still seeing 21v on the +18 and 51v on the phantom without that chip?  Does this give you any more clues as to what's going on here?

Could the BD237's be bad?  Is there any way to test those transistors?  I'm assuming the one in the -18v side is fine, but it's possible the + side is bad?  Just a thought...

EDIT - I actually JUST noticed that someone has changed the +18v transistor to an NTE184 - not sure if this is compatible or not, but it's not the part that was specified in the parts list? 

Looking at that NTE184, left to right, I get 22.3v / 37.2v / 22.3v - this seems unusual compared to the transistors I've measured before...a clue?

 

[radardoug]

The phantom is completely separate from the +18. NO connection between them.
Your phantom circuit is now close, the different 7824 might be the problem. They have made this circuit using a 24 volt regulator and jacking up its ground reference. The current in R11,R12 might be wrong enough to give you the wrong voltage.
Put a 10K trimpot in instead of the 3.9K, adjust for +48 volts, then measure the trimpot value and replace with a resistor.

On the 18 volt side, with the 723 out, take out the first transistor nearest the 723. Does voltage out drop to zero? If not, that transistor is faulty. Transistors can be measured with a voltmeter with a diode range. The transistor should measure 400 to 600 mV in the forward conducting direction, and open with the terminals reversed, across the base-emitter junction and the base-collector junction.
Across the collecter-emitter junction the device should be open both ways. Note that an ohmeter cannor be reliably used to measure transistors, but will indicate a shorted transistor. Transistors can fail to short or to open circuit.

The purpose of the transistors is to increase the current that the supply can produce. The 723 from memory will only regulate at the 100 mA level. This supply needs to produce several amps.
You could replace both transistors with a power darlington, the transistors are in a darlington configuration.

 

[Siegfried Meier]

Ok, I'll try this.

But, the question remains.  Why does the phantom voltage change when I put in the newer style 723 chip if they're not in any way connected?  I had an increase of over 12v with the newer voltage regulator...

Thanks for your help!
Sig

 

[mjrippe]

And Doug, isn't the +18v rail being used to elevate the ground of the +48?  So why would you say they are not connected?  And how do they get the extra 6v (18+24=42)?

 

[radardoug]

Ah yes, I hadn't picked that up! Michael Mouse indeed! OK, so work on the 18 volt supply first and get that going, then work on the +48.
That means that the 24 volt regulator is jacked to produce 30 volts only.

 

[mjrippe]

Sig, going back to the +18v side, I thought both transistors T1 and T2 were 3055 (parts list is cut off) which is why I told you to replace both 3055.  Probably that NTE replacement should be replaced as well   Also check the resistors on either side of your trim pot.  It looks like once you get the +18 settled the +48 will follow.

 

[Siegfried Meier]

Excellent!  This was my thought as well.

Yes, at first I thought both 2N3055's were together, but there's one for each rail + & -18...I think that the NTE is at fault as well.  I'll order one (or several..) up and see what happens.  Appears those resistors might be ok, I gave them a quick check, but we'll see what happens when we replace the transistor.

Thanks!
Sig

 

[Siegfried Meier]

That did it!!  We are back up and running, perfectly.

Now, since I've never actually been able to use the console yet (we got it in late 2011, and the PSU blew the second I fired it up without the console even connected), I'm a little leery of plugging it in.  The same guy who did the work on this said he blew something in the master section, and he thinks it was because of him replacing some of the chips.  I have all the old ICs still, so I think my best best would be to try to replace all those again back to the originals, or at least fire the console up with no master section connected.  Any advice moving forward?

Thanks so much guys. 
Sig

 

[mjrippe]

Either no master, or only the master section if you're feeling brave and have spare parts for the PSU now!  You'll have to try it at some point, so perhaps go back to the stock chips and throw the switch 

 

[Siegfried Meier]

Ya, I have a sneaking suspicion that the reason this PSU blew up so bad is because of the changes he made to the master section, which is not advised by Duco (owner and founder of D&R - it's actually pretty cool that he still replies with info on this desk, and they even carry a limited supply of parts for them still, after nearly 30 yrs!).  I saw posts of the guy who owned it before online asking questions about modding things as well as the PSU.  After he made a mess of it, he decided to get rid of the desk.  I'll poke around in there first and see what looks new, and stick the old ones back in place.  Then, I will eventually, as you say, have to throw the switch and hope for the best.

Thanks again!
Sig

 

[radardoug]

Although it is a lot of work, if you are worried that there could be faults in the desk, power it up in sections. Probably means pulling all the modules out.
Start with just the master section on. This also helps the supply, in that it is not trying to supply full current all at once. Also you could put temporary fuses in the supply at say 1 amp. Or use a bench supply with controllable voltage and current.

 

[mjrippe]

Yes, Doug is right again!  You don't want to make 5 amps available if you are only powering the master section.  By the time the fuse blows you might have a charred board.

 

[Siegfried Meier]

Ok!  Have made some major progress with the console.  We're close, but now I'm finding my 2N3055's are getting really hot on the PSU when it's all fired up - even the heatsinks and chassis is so hot I cannot touch it.

I've completely recapped the large caps in the PSU now (couldn't find Axials that worked, so we've made Radials work with jumper cables).  Its still getting hot though unfortunately, but I took measurements of the caps and here's what I've found - I've numbered the caps so you can see which ones I'm talking about (what I believe is the +48 side, with the 3 caps together, I've named that 5):

With respect to ground, here's what I get at both the positive and negative terminals of each cap:

1) +15.8 and -18.2
2) +15.8 and -18.2
3) +33.6 and 0 on the negative terminal (this seems odd to me)
4) 33.6 and 0 on the negative terminal (same, also seems weird)
5) +69.2 on the positive terminal and +18.2 on the negative terminal (which I also found odd)

Despite these readings, everything does seem to work, including phantom power, on the console.

Any ideas on those strange 0v voltages I'm getting?

 

[mjrippe]

Hi Sig,

Are you SURE you are measuring from Ground?  It looks more like you are measuring from one of the voltage rails.  Perhaps if you re-label the caps according to the schematic it will be more clear for us.

Mike

 

[Siegfried Meier]

Hmmm well I'm measuring from the ground of the chassis - ie. screws that attach to the chassis?

 

[mjrippe]

Ok, measuring from those same screws, what do the voltage rails measure?  Something is definitely odd, but the numbers kind of make sense.  15v + 18v = 33v for example.  And we know the +48 supply is connected to the +18v supply (18 + 48 = 66 so 69.2 is not too odd).  Again, having the schematic part numbers instead of arbitrary ones would help.

 

[Siegfried Meier]

Ya, unfortunately the boards don't say anything about which cap is which, so it's hard to guess from the schematic...

I did get this message from Duco this morning though, which was kind of a surprise:

From what I see there is no original power transformer mounted.
Therefore +33.6 volt DC bringing down to +18 volt produces a lot of heat dissipation you encountered already by burning you fingertips.
Attached is the circuit diagram of the power supply.
The transformer should give 2x 21volts AC and 48volt AC at 300VA
Resulting in 2x 30 volts DC and 1x 69 volts DC prior to the voltage regulators.

We still have these transformers


Kind regards |  Vriendelijke groet
Duco de Rijk

 

[Siegfried Meier]

Hmmmmm I think Duco is trying to replace my transformer with a single one - this PSU has 2 inside of it, one for the mains and one for the +48v I think.  It definitely appears to be heavily modded, and in speaking with a previous owner, he said they replaced all the opamps which tells me right away that they needed to beef up the supply in order to make them work.

So, maybe I'm exaggerating this heat thing?  I mean, yeah, it's hot.  But, it's entirely possible that it was also hot for all those years that they used it?  Our TAC supplies have fans in them for this exact reason.  What if I attached another heat sink to the existing sink to dissipate more heat?

The console works beautifully right now - I got all but 2 channels back in working shape, and the master section (which was totally dead) is now 100% working.  I just don't really know what else to do, and tearing it all apart to put another trafo in seems like I'm heading in the wrong direction...

Any thoughts?
Thanks guys,
Sig

 

[mjrippe]

Well, there is an old saying - "If it ain't broke, don't fix it"  ;D