Zener with Pass Transistor Regulator Circuit not Giving Expected Voltages

[MainePicker]

Hi All,

The attached circuit schematic is to the power amp section of a guitar amplifier I own. The parts of the circuit at Q1 and Q11, and their associated zeners, should supply the +15vdc and -15vdc. However, I get +15 and -13.

I can "trick" the negative supply up to -15vdc by adding 1n400x diodes in series with the zener and things will work fine. However, I'd really like to know why it's not working as expected.

The power amp did fail in the past, and many components associated with, and along, the -15vdc supply were damaged. All have been replaced. Every electrolytic cap in the amp has been replaced. I replaced IC1 at one point, just to make sure it wasn't dragging down the -15 supply.

Q1 and Q11 were upgraded to TIP29A and TIP30A which handle higher voltages. The output transistors have been changed. I'm wondering if the new transistors could somehow account for the low voltage on the -15 supply. But, if so, then why wouldn't the +15 also be affected?

My question is this: What types of things should I be looking at? Are there any likely suspects in the circuit that could be dragging down the -15 supply?

The +/- 40vdc rails are fine and balanced. There's plenty of supply voltage getting to the zener.

I'm scratching my head at what could be causing this.

Any thoughts/ideas/help/suggestion?

Any questions, just ask. I'll be happy to supply any additional info needed.

Thanks, all!

[Ian.M]

There's really not a lot loading the -15V rail apart from the BC107 input stage/phase splitter and the dual OPAMP.  You can calculate the load current on the + and - 15V rails if you measure the voltage drop across R3 and R25.   

Check R23 and R25 haven't increased in value.  Swap the Zeners and see if the fault moves to the +15V rail. 

[innkeeper]

VBE on a tio30 is 1.3v so the voltage id expect to read is -13.7V
The diode "trick" is perfectly viable or find yourself  a zener close to 16.3v

[Ian.M]

VBE on a tio30 is 1.3v so the voltage id expect to read is -13.7V
The diode "trick" is perfectly viable or find yourself  a zener close to 16.3v

Where do you get that idea from?  The schematic has R25 as 125R with 10V across it so Ic is 80mA.  If you consult fig.2 of the Fairchild TIP30A datasheet, Vbe_sat is under 0.8V @ 80mA Ic, but as it isn't saturated it will be less, probably around 0.65V

Find the fault and fix the problem, don't just patch the symptoms. 

[IanMacdonald]

Both of the regulator transistors have a collector resistor, so measuring the drop across these and applying Ohm's Law would give some idea of the current draw.

If putting a diode in series with the zener raises the emitter voltage then that shows the zener is conducting and the transistor cannot be saturated.  Therefore, probably just an out of tolerance zener.

As pointed out you will only get 14.4 or so anyway without the extra diode, because of the Vbe drop.

[MainePicker]

Thanks, guys. Good info. I will calculate the current next time I have time to work on the amp. Some other info based on your input:

1.) R25 *is* at the top end of its tolerance @ 132 ohms.
2.) The zener has been replaced and I got the exact same voltage
3.) When I did temporarily trick the circuit with 2 series diodes, it brought the voltage to -14.9 and both the resistor and transistor ran much cooler.

Not sure what all that means, but I figured you might might know ;-)

Thanks for the help. The amp has preamp problems, too, but I don't even want to get into them until I feel secure that the power amp section is working properly.

Peace,
Bud

[Ian.M]

What voltage drop are you measuring across R25?
Is the preamp loading the -15V rail?
What's the resistance of R23?   If it's gone high resistance, it would explain a lot.
Can you check the H_FE of Q11?
Was C7 replaced with the correct polarity?
Have you scoped the -40V rail to check for excessive ripple?

Edit:
Please measure all the voltages around Q11 and post them to 2 d.p . if you are using a DMM or 1 d.p. if using and analog meter.

[innkeeper]

Where do you get that idea from? 

I think you my have mis-read the TIP30A spec. The sat spec is usualyfor Vce and thats what i see when i look at it.
The Vce(sat) has no coloration the regulationon this circuit, but the on voltage of a bipolar transistor, the VBE(on) does.
that voltage will be seen as an additive or subtraction (depending in your circuit) voltage in series with your zener.
So, VBE is the voltage we need to consider here.
http://www.mouser.com/ds/2/68/tip30-845497.pdf
http://www.onsemi.com/pub/Collateral/TIP29B-D.PDF

VBE 1.3 Max (so yes it can be less then 1.3 - best to measure it)

Thanks, guys. Good info. I will calculate the current next time I have time to work on the amp. Some other info based on your input:

1.) R25 *is* at the top end of its tolerance @ 132 ohms.
2.) The zener has been replaced and I got the exact same voltage
3.) When I did temporarily trick the circuit with 2 series diodes, it brought the voltage to -14.9 and both the resistor and transistor ran much cooler.

Not sure what all that means, but I figured you might might know ;-)

Thanks for the help. The amp has preamp problems, too, but I don't even want to get into them until I feel secure that the power amp section is working properly.

Peace,
Bud

The resistor being cooler could simply be that with the voltage rails unbalanced the amplifier circuit could have been conducting with no signal input.

The 4558 op amp that used to bias Q2 and the whole dc coupled amp circuit is driven from the +/-15v.  if the +/- 15v rails that power it are not balanced, the bias could be off and one side of your amp would conduct.
looking at the circuit id not be surprised if 1/2 of your power transistors were also getting warm if you had the speaker connected.

I happen to be currently working on a tip122 / tip127 +/- voltage regulator circuit myself with a VBE of 2.5 so it threw me for a second also.

Don't take my word on VBE being part of the equation for your output voltage. .....Measure it yourself. ... measure the actual VBE from Base to Emitter, and the actual voltage across the zener. (or across the 2 diodes and zener if you ahve the 2 diodes in)
Subtract the two and i bet you end up with your output voltage.

This isn't to say there might not be something wrong. Check for excessive idle current with the +/- 15v balanced  (close to the same voltage + and -)

measure the voltage across r25 and do some math, see what the current draw is.
Basically though, if the voltage  is significantly higher the the expected 10v, then our circuit is drawing more current then it should and you should trouble shoot that.

My belief is still threes not a darn thing wrong. the circuit is working as it should.  If so, then leave in your two 1n400x diodes and call it a day or get yourself an appropriately rated voltage zener that takes the VBE into account

they make zeners in 16.1,16.2,16.3 voltage ranges just for this reason.
and for those with VBE of .6 they have 15.6v zeners ... i'm sure you have seen that value, it is popular.

this is where getting out your meter and actually measuring will tell you what you need.

however ...... (diodes)6v+.6v = 1.2v   
15v+1.2v= 16.2 (total zener voltage)
-16.2v + 1.3(Vbe) = -14.9V

Curious how the math works out

Good luck, keep us posted

 (edit for grimmer)

[Ian.M]

Note the conditions for that Vbe(on) you were quoting:

    VBE(ON) VCE=4.0V, IC=1.0A .....  max 1.3V

As the schematic resistances and indicated voltages tell us the load on the -15V regulator *should* be 80 mA, its fairly obvious that the Vbe will be close to the typical BJT 0.7V as the TIP30A is silicon and not a Darlington.   15V Zener^* - 0.7v = 14.3V out.   The O.P.'s 13V out indicates a problem . . . .

I referred to the Fairchild TIP30A datasheet (linked in reply #3) *BECAUSE* *IT* *HAS* *GRAPHS*.   Graphs in a datasheet often let you tease out more info about the *TYPICAL* behaviour of a device, and in this case, as the graph "Figure 2. Base-Emitter Saturation Voltage Collector-Emitter Saturation Voltage" has the base voltage with a current equal to Ic/10 applied to the base, (far more than in the O.P's circuit), we can be reasonably certain that Vbe @80mA Ic  will be less than the indicated Vbe_sat @80mA Ic_sat of just under 0.8V. 

So IMHO the maths indicates there's at least 1V less out than the O.P. *should* be getting.   

That's why I have asked for more measurements . . . .

* A 15V Zener will only deliver 15V if it is right in the middle of its tolerance range and passing its nominal current.

[MainePicker]

Thank you, Ian. I've taken these readings already and recorded them in my notes. Here's what I have:

>>What voltage drop are you measuring across R25?
R25 Vdrop: 4.22


>>Is the preamp loading the -15V rail?
No. It's disconnected for now, till I get the PA problem solved.


>>What's the resistance of R23?   If it's gone high resistance, it would explain a lot.
Don't remember the exact measurement, but remember that it was more in tolerance range than was R25 so I didn't worry about it. Though I did wonder about R25 being up at 132 ohms, pretty much at the edge of its 10% rating.

>>Can you check the HFE of Q11?
I just measured the one I pulled out and its 251. I could put it back in. But I suspect the one I put in is the same. I'd swapped it out "just to see" and the -13vdc remained. I also swapped out the zener (again) and no change.


>>Was C7 replaced with the correct polarity?
Yes.

>>Have you scoped the -40V rail to check for excessive ripple?
Yes. Didn't really find anything excessive, but replaced those anyway with 3900uF low-esr Nichicons and no dice :-/


>>Please measure all the voltages around Q11 and post them to 2 d.p . if you are using a DMM or 1 d.p. if using and analog meter.
CR5 Vdrop: 13.58
R25 Vdrop: 4.22
R23 Vdrop: 17.2 (17.1 to 17.2 a bit jumpy)
Q11-E   -40.70
Q11-C   -40.71
Q11-B   -13.61
Q2-E           -0.60 (test point on the schematic -- right on the nose)

One interesting thing to note, which you might have noticed already, is that the voltage drop across the collector resistor R25 is not 10vdc. I had discovered one or two anomalies in the schematic. Another example is that the schematic has the rails at +/-45 in some places and +/-40 in others. There is also a discrepancy with the labeling of Q6 and Q16 -- they are crossed up with the PCB silkscreen. So when I saw the +30 notation near Q1 on the +15vdc regulator circuit, I though maybe the -30vdc notation had been hastily added to the negative side by whomever was making the notes on the schematic. But the two circuits are not mirror images. So this made me think the "-30VDC" notation might be another anomaly in the schematic.

But since we're revisiting it . . . do you think it *should* be a 10vdc drop? Seems very far off from what I'm measuring, and the -15 supply is only off by a volt or two. So it doesn't seem to add up either way.

FYI, all the AC and DC "test points" on the schematic check out fine. The only things off are the implied test-points (such as "-30VDC") around Q11.

TP2 was a little high, reading .895Vp-p, which is also what TP3 reads. Small voltage, but significant percentage, so I dunno. I kind of discounted it because everything else checked out.

Thanks for all the help. Hopefully I haven't added more confusion . . .

Peace,
Bud

[Ian.M]

>>Please measure all the voltages around Q11 and post them to 2 d.p . if you are using a DMM or 1 d.p. if using and analog meter.
CR5 Vdrop: 13.58
R25 Vdrop: 4.22
R23 Vdrop: 17.2 (17.1 to 17.2 a bit jumpy)
Q11-E   -40.70 
Q11-C   -40.71
Q11-B   -13.61
Q2-E           -0.60 (test point on the schematic -- right on the nose)

<snip>

But since we're revisiting it . . . do you think it *should* be a 10vdc drop? Seems very far off from what I'm measuring, and the -15 supply is only off by a volt or two. So it doesn't seem to add up either way.

FYI, all the AC and DC "test points" on the schematic check out fine. The only things off are the implied test-points (such as "-30VDC") around Q11.

Please double check the reading in red!    Better yet, measure the Vbe of Q11  directly.

So the load on the -15V rail with no preamp is currently 32mA.  If the preamp draws just under 50mA, that would be consistent with the 80mA expected from the schematic's indication of a 10V drop across a R25 of 125R

-13.61V on the base indicates the Zener is either being heavily loaded or is current staved, or is out of spec, *OR* possibly there is a cracked track or dry joint between the Zener and the transistor base adding about 10K of unwanted resistance.

[innkeeper]

What ever became of this repair... any updates?

[MainePicker]

What ever became of this repair... any updates?

Please accept my apologies. Hopefully I will get back to this thread pretty soon. What happened is, in the midst of this, my main amp that I use for rehearsing and gigging SMOKED one night. Repairing that one became a priority. I got it going, but during the repair I damaged one of the pots when I pulled off the knob and the shaft came with it. So I'm waiting on that part before I completely put the amp back together.

Thankfully I got it going, though, and once it's officially off the bench I'll get back to the other amp. I do intend to answer all of the above questions. My main amp was a priority though, and I'm in a small condo can can only have one project going at a time.

I will have an update soon!

Thanks for hanging in there.

Bud

[innkeeper]

no problem, i was just curious how you making out with it.
i know what that is like though, just in the past month i was fixing and upgrading / modifying a DBX 286A when my HP DSA broke,  so ended up repairing the test equipment first
crap happens..

[MainePicker]

no problem, i was just curious how you making out with it.
i know what that is like though, just in the past month i was fixing and upgrading / modifying a DBX 286A when my HP DSA broke,  so ended up repairing the test equipment first
crap happens..

A day or two before the Marshall smoked, I was thinking about how old it was, and how it had never blown up on me before. Then I thought, "Uh, don't think that." Too late. Within two days it smoked!