-10v dc - dc converter repair advice needed

[rnie]

Hello,

I have a broken drum-machine here that seems to work fine other than that i’m not getting any sound from it.

So i opened it up and found that there is no -10v present. thus far i removed and checked Q1, Q2 (npn), C8 and D3(zener) in the -10v dc - dc converter but all those seem fine. This leaves me with the little custom sumida coil/inductor/transformer T1.

There is no stock (or datasheet) for custom roland parts so i'm left with finding a similar part. To be honest i dont know its purpose so that makes it even harder to find a replacement if such a part exist.

Another option would be to bypass the dc - dc converter with a separate module, PTN78060A or a MAX765 based module or something from aliexpress. If i go this route I dont know how to implement the line after R2 (i-20 to IC30A)!

http://www.synfo.nl/servicemanuals/Roland/TR-707_SERVICE_NOTES.pdf (page 10 top left corner)

Thanks allot, Any advice is much appreciated!



Edit: I just isolated the -10v supply from the rest of the board and the readings are a bit different (+0.7v is 0v now). But the problem persists.

[floobydust]

You have no enable voltage for the oscillator. Before you blame the transformer, see why R2 has no +ve voltage coming from IC30A. That is why Q1 is staying off.

I think the Gate Array enables this DC-DC as part of the start-up sequence for power. Roland must want the -ve rail to come up after the +ve regulator and MCU initialization, so IC's don't get damaged due to reverse flow from between -ve rail to the +ve rail, or maybe thump protection etc.

I would lift one end of R2 and with the unit on (+5V present) connect 4.7k from Q1-B to +5V. The DC-DC should run OK.
Unless it's overloaded due to a short on the -10V rail. But I think the MCU should be switching the DC-DC on.

This is all wrong,  MCU IC30A provides pulses to the two transistors of the DC-DC converter. It is not a self-oscillating.

[rnie]

Thank you, that makes allot of sense. I'll give it a go.

is IC30A part of IC30? Edit: I see it now!

[rnie]

I would lift one end of R2 and with the unit on (+5V present) connect 4.7k from Q1-B to +5V. The DC-DC should run OK.
Unless it's overloaded due to a short on the -10V rail. But I think the MCU should be switching the DC-DC on.

I gave it a go and somehow fried R49  I replaced it and all seems to be like before (i hope!!).

I connected a 4.7k resistor to the base of Q1 powered the unit on then connecter the other side of the 4.7k resistor to the +5v line.
  i dont think i shorted something else doing so or?

[rnie]

It was a broken trace along that  voltage enable line! the unit seems to work fine now! thanks allot for pointing me to it!

[floobydust]

It's great you got it going, TR707 is a blast! TR727 a little too Latin American but a few sounds not too bad.

[rnie]

Yes very happy! I had a perfect one (probably never used) but sold it with immediate regret.

So i could buy this one for cheap.. i think someone stood on it because the case is cracked at multiple places.. probably should have expected a damaged pcb.. but well eventually i found it . (with help.. thanks so much! ).

[Graboid]

Hi all I am in the exact same boat as rnie but I do have the ve from IC30A (at least it seems so) but no -10V rail.  As it stands, the 707 starts up fine and all other functions such as programming and buttons respond fine, just no sound at all.  It would seem like the -10v to the audio board is not getting there.  I am still trying to wrap my head around negative voltage, so any help getting this thing making noise again would be helpful.  At the moment, I am eyeballing Q1 and T1 as the culprits.  Thanks   Hope someone more knowledgeable than myself can lend a hand.

[floobydust]

Either the DC-DC is having a problem or it is overloaded.  I suggest measuring some voltages with a multi-meter; the -10V output and Q2 collector (T1 pin 4).

Gate array IC30 provides pulses to the two transistors. So Q1, Q2, D2, C8 could have failed.

The transformer is fragile, don't bump it. OP had an open PCB trace, so maybe look at the soldering and traces for damage.

The -10V rail goes to many components. To troubleshoot a short on the -10V rail, you would have to measure the voltage drop across DC supply resistors R145, R130, R118, R135 etc. and see who is hogging current, in the case of a shorted op-amp or something.
TR-707, TR-727 schematic, service manual notes

[Graboid]

Thank you very much.   I will probe the areas you have suggested and see if I get any closer to the culprit.  I totally hear ya about being gentle with the transformer.  That lil beast looks like it would be difficult to source.  I will check things and report back if I get any notable readings.  I did already replace D2 as it was just acting squirrely with readings.  Huge relief to get this project going again!

[Graboid]

Alrighty so I ended up swapping out C8 as it was bulgy and did some probing around as you suggested (i think).  So now Q1 is showing 11.17 and D2 is still showing -0.69V.  Q2 is showing 0V so maybe it is dead or not being fed?  It seemed like it was working before but I curse my old and feeble brain as I am not sure now.  Is there anything glaring that I should check/do?  Thank you for your time and patience.  This power system is not clicking in my brain for some reason    So is the ve for the Qs to oscillate and keep the caps charging and discharging and keep -10V at the plate or am I just way off?

[floobydust]

Q2 was 0V at the collector, or the base?

I would measure the voltage on either side of R2 to see if the big IC is sending any pulses out. You need some positive voltage pulses at the base of Q1 to get any output from the DC-DC converter.

OP had an open PCB trace, the long run from R2 to IC30.

How it works - Gate array IC30 outputs square-wave pulses to Q1, who then turns on (pulses) Q2 which pulses the transformer. D2 is the output rectifier.
Once the output reaches -10V, zener D3 and D18 shut off Q1/Q2 as needed to maintain -10V, for regulation.

[Graboid]

Thank you for the explanation it makes more sense.  I had the Q's doing oscillation instead of amplification.
I checked and checked all sorts of connections between points and haven't found any breaks in continuity yet.  I even borrowed my wife's reading glasses and got down in there close but just did not find anything like that.  The board is in relatively good shape minus the previous owner's repair job that was backwards. 

Okay, so tested things again and these were the readings in response to your questions  
Q2 was 0V at all 3 pins. 
Q1 showed 11.17V at the base and collector(?)
R2 was 11.17V on the side closest to the transformer.  The other side was 5.65V.   It rang true on continuity to IC30a.
Maybe something will stand out to you.  I am very thankful for your time and patience.   

[floobydust]

I think your two transistors have failed. I would pull and test both of them.
Q1 has C-B short and E-B open; R2 should not have 11V on it.
Q2 with 0V on all pins, it should have some voltage on the C, unless the transformer primary is open-circuit. Check ohms on T1 from pin 3 to pin 4.

[Graboid]

Cool.  So I tested T1 pins 3&4 to the best of my ability (there are a lot of random pins there and unlabeled so I rang out to the Qs they were allegedly attached to haha).  Everything rang to each other.  Like in continuity mode/200R everything rang out as connected.  This is on the board though so maybe other connections are causing that.  When I put it up to 2K (my other lowest setting on dmm) it all measured 0.0 or 0.1ohms.  I probably did something wrong.  Sorry.
I will pull Q1&2 tonight and test them out.  I think my other meter has holes for em.  They are both NPN, correct?
I am getting that sinking feeling if that oddball T1 is toast.  I have seen them before in my random tinkerings in that form factor but the more common box type is what I usually see.  Thanks for putting up with this. 
I am off to de-tar some paving equipment but only half day so should be home early and can hopefully get on those Qs

[Graboid]

Alrighty long weekend.  Cat passed away.  Got the Qs pulled and tested.  Q1 isn't behaving correctly and Q2 is complete toast.  I am trying to find replacements online but am not having very good luck so far but am a bit distracted anyway.  I called Roland US and they had none as to be expected.  Thanks for your help, Floobydust! I feel sure these are the problem.  If you have a source for these that would be cool.  Otherwise, not gonna hassle ya again until I get them here and installed.  Hopefully will have good news. Peace and thanks again

[Graboid]

Well that didn't go well.  They sent me a bunch of Vietnam era transistors.  If you ever see this, Floobydust, please let me know of some alternatives for Q1 and Q2.  I am another 40 bucks in the hole with a bunch of useless transistors.  I would love to just get this thing done and at least see if that gives us the -10V rail without hopefully having to chase down that damn weird ass transformer.

[floobydust]

Q1 can be any vanilla NPN with JEITA (Japanese) E-C-B pinout, like 2SC1815 or KSC1815. Or BC546 good but C-B-E pinout, so you'd have to bend the leads.

Q2 is 2SC1384? It has medium power, high current (1-1.5A) rating so I would sub KSC2690 as a tougher replacement, although a slightly different package.

If Q2 shorted C-E then I would think the transformer would get roasted, smoke and that kind of thing. So the fuse here... was either the old Q2 or the transformer primary winding going open-circuit. If you get continuity on transformer 3-4 then it's worth a try, I'd keep going.

[Graboid]

This is simply tremendous, floobydust!  I cannot thank you enough!  Exactly the info I needed to get things back on track and explained in a way that even I can understand As you can tell I am a bit gunshy when it comes to swapping PSU section parts willy nilly.  After reading your explanation, I feel like my gut was right on a couple of things so I have a little more confidence in the pocket. I am going to go thru my parts as I may have a couple of the ones you mentioned and order the other.  I am so excited, not even worried about the transformer at this point.  It isn't ringing open on the board with the Q1 and Q2 gone, so we may be okay (fingers crossed, knock on wood).  Please pat yourself on the back and know that you have made at least one person'd day very much better

[Graboid]

Well it has been a long wait but got the correct transistors for Q1 and Q2.  Got those swapped no problem and powered it up and all works like before but still no -10V      At this point I am gonna start hunting down that transformer.  I really wish I had just gone ahead and done that but oh well.  All of the voltage measurements were the same as before.  I guess at least I didn't kill anything lol.  I was really certain it was those transistors as one was fried and the other was squirrely but I guess the transformer was gone too.  All of the pins were ringing out to each other so something seems off to say the least.  11.97V goes in but nothing comes out beyond .67V

[Graboid]

Arrrgh.  Looks like that is the end of the line.  I see others from years ago looking for the same thing but it does not exist beyond a dead board.  The only one on reverb right now has it pulled.     So insanely frustrating.  I will wait a couple of days to see if I get any response and then it is time to just smash the hell outta this thing.   That at least will provide some sort of pleasure from this headache lol.  Thanks again for all your help, Floobydust.  I could not have gotten this far without ya.  Sorry we couldn't save it.

[floobydust]

You give up too soon, the world needs more cowbell lol.
You can do surgery on the transformer or just put in a little DC-DC converter module to make the -10V.

I would get a module with 12V input, and +/-5V out (series to get 10V) or a 12V output if you add a voltage regulator IC to get down to 10V.

Like Meanwell SPR01M-12 or CUI PDM1-S12-D5 or Murata NFM.

It would not be much to make a small PCB that fits the DC-DC converter and a voltage regulator IC. That is what I would do.
If this transformer/transistor short is a common fail on the TR-707/TR-727 then it might be worth it.

[Graboid]

I appreciate the effort, floobydust.  I am just gonna try it with an atx psu I have here that provides the voltages I need once I put a few diodes on the -12V rail to dumb it down to -10 and attach the 5v and gnd.  That will hopefully let me see and hear if anything on that rail even works anymore.  I also appreciate the part suggestions but the whole negative voltage thing just messes my dang head up.  I am pretty crestfallen at this point and suck at power supply stuff.  Will try the pseudo benchtop psu and let you know how it goes.  I doubt I could build something that I would consider safe at this point for human or machine.  Take care and hope you have a good weekend and thanks again for all of your help.

[Graboid]

Alrighty I got my makeshift PSU connected using the gnd 5v and -12v rails.  I managed to get it (up) down to -10.87v so it was in fudging voltage range.  Fired it up and the main logic functions and buttons were working and lo and behold I heard the static crackling of a drum sound coming thru my amp, which has never happened before so I guess there is some life in the old girl yet!   I was nervous of burning something out so I kept the tests short and sweet.  Not sure where to go from here.  I would be willing to purchase the parts you suggest if you wouldn't mind walking me thru some of the ideas behind what we are doing.  I am kinda teetering towards just taking it into a shop to be repaired but if they can't get the transformer either I would like to watch them and learn what was happening and what I missed.  The major hurdle there is that I am currently a broke loser and balk at the idea of paying hundreds of dollars in bench fees just to solder in a couple of parts (the eternal conundrum haha).  Floobydust, what are your feelings on this?  I trust your opinion good or bad and appreciate the time and effort you have already put towards this.  I would happily kick in a donation to you for your time if I can get her cowbellin' again but totally understand if this seems over my head.  Thanks

[microbored]

Reviving this old thread.  My TR-727 T1 (S097744 DC-DC converter) is toast!!!

Is this what floobydust is suggesting on how how to build a replacement using a -5V DC-DC part (PDM1-S12-D5-S) in series?

https://www.digikey.com/product-detail/en/cui-inc/PDM1-S12-D5-S/102-2712-ND/4009637

Thanks!

[fzabkar]

I believe the idea is to use a single "PDM1-S12-D5-S" module. According to the datasheet, the module's input and output are electrically isolated, so this modification should be feasible.

Connect the Vin and Ground pins on the input side of the module to 12V and ground on the PCB. Then connect the -5V pin of the module to the -10V rail of the PCB, and the +5V pin of the module to ground on the PCB. This assumes that the -10V components on the PCB draw less than 100mA (the rated current of the module).

https://www.cui.com/product/resource/pdm1-s.pdf

[microbored]

Ahh, I understand now now!!!  Any thoughts why he suggested to use a 12V output with voltage regulator IC instead?  Thanks!

[motosega]

i made a replacement negative psu module for the tr707, it uses icl7660 negative voltage converters.

https://synth.ewancolsell.com/home/tr707-negative-vibes-10v-psu-replacement