Help please/ Debugging how to?

M

Moltenmetalburn

Member
Ive got a pile of non working builds piling up. It seems for every one I get built another doesn’t work and I don’t know enough to get through them on my own. after an integral preamp I built today doesn’t work yet again I am pretty frustrated.

My integral is farting and distorting. I reflowed everything thrice. Then I swapped all the Electrolytics, one film I suspected, and the Q2 transistor. Still a farting crackling mess.

After blindy swapping 10 parts and no change I need help.

How do I learn to debug and use an audio probe??? I bought one but haven’t experimented yet.

getting so frustrated it makes me want to give this up at times.

I see some extremely talented people here helping others debug, I could use your help as well.

thanks!
 
Sort by date Sort by votes
GenoBluzGtr said:
a 550 should be fully interchangeable with the 548 (and 5465, 547, 549, as well), but there are some differences... I don't think they would cause the problem you're experiencing. There is something causing Q2 to clip/distort/saturate. I'm still not sure about that 1M resistor.. when he used the 1M, did he compensate elsewhere in the circuit? That's almost a HUGE reduction in the amount of voltage drop, unless something else was changed to keep the ratio the same. Maybe I'm off base, but that bugs me and it's in the right spot to cause an issue.
Click to expand...
I don’t know. Hopefully he will chime in at some point, but it does work fine in the other build.

this board has a built in charge pump which is how it reaches the higher voltages I mentioned earlier from a 9V supply.
 
Upvote 0 Downvote
Moltenmetalburn said:
Charge pump:
8.5 8.50
3.8 6.20
0.0. 4.00
0.0. 0.00

opamp:
0.00 0.00
24.9 28.2
24.9 25.0
0.00 0.00
Click to expand...

Don't make us guess which is pin 1. I'm going to assume that this is viewed from the top and that the notch is at top for both.
Charge Pump (IC100)
The only pins that will have meaningful DC readings are pins 1, 3 & 8. They all look ok.

Opamp (IC1)
pins 2 & 3 should be equal and around 25V and they are, so that's good. Pin 7 is at Vcc, that measures 28.2V, so that's good. Pin 4 is ground. Pin 3 is Vref, measures 24.9V so that's good. Pin 6 should be around 17V and it's not. What could make that happen? R5 & R6 are the right values. Q2 seems to be working. Measure the voltage across R6: black meter probe on the end closest to IC1, red meter probe on the other. If the circuit was working right, we'd see about 6.6V. Another possibility: C2 might be shorted.

Geno,
The collector current is Q2 is (Vcc-Vref) / R5 = 33uA. HFE is in the 100's when Ic = 2mA. HFE decreases at low Ic, but let's just assume it's 100. That means the base current is 330nA. The voltage drop in a 1M resistor is 330mV. The voltage drop in a 6.8M resistor would be 2.24V. That would cause the emitter voltage to be about 2V lower compared to a 1M resistor and consequently, IC1 pin 7 would also be about 2V lower. Either way, the circuit works fine.
 
Last edited:
Upvote 0 Downvote
Thanks for the explanation Chuck, It makes sense when you show the math, but on the surface it would seem to impact the performance of Q2, I guess just not as much as I was thinking.

I like the troubleshooting too, that makes perfect sense that either R6 is off or C2 is not blocking the DC. Either of those would cause the symptoms.
 
Last edited:
Upvote 0 Downvote
Chuck D. Bones said:
Don't make us guess which is pin 1. I'm going to assume that this is viewed from the top and that the notch is at top for both.
Charge Pump (IC100)
The only pins that will have meaningful DC readings are pins 1, 3 & 8. They all look ok.

Opamp (IC1)
pins 2 & 3 should be equal and around 25V and they are, so that's good. Pin 7 is at Vcc, that measures 28.2V, so that's good. Pin 4 is ground. Pin 3 is Vref, measures 24.9V so that's good. Pin 6 should be around 17V and it's not. What could make that happen? R5 & R6 are the right values. Q2 seems to be working. Measure the voltage across R6: black meter probe on the end closest to IC1, red meter probe on the other. If the circuit was working right, we'd see about 6.6V. Another possibility: C2 might be shorted.

Geno,
The collector current is Q2 is (Vcc-Vref) / R5 = 33uA. HFE is in the 100's when Ic = 2mA. HFE decreases at low Ic, but let's just assume it's 100. That means the base current is 330nA. The voltage drop in a 1M resistor is 330mV. The voltage drop in a 6.8M resistor would be 2.24V. That would cause the emitter voltage to be about 2V lower compared to a 1M resistor and consequently, IC1 pin 7 would also be about 2V lower. Either way, the circuit works fine.
Click to expand...
Ok, noted. I edited the post for clarity in respect to the pinout. ??

I will get to this and report back my findings, thank you for your help again. ?
 
Upvote 0 Downvote
Chuck D. Bones said:
Don't make us guess which is pin 1. I'm going to assume that this is viewed from the top and that the notch is at top for both.
Charge Pump (IC100)
The only pins that will have meaningful DC readings are pins 1, 3 & 8. They all look ok.

Opamp (IC1)
pins 2 & 3 should be equal and around 25V and they are, so that's good. Pin 7 is at Vcc, that measures 28.2V, so that's good. Pin 4 is ground. Pin 3 is Vref, measures 24.9V so that's good. Pin 6 should be around 17V and it's not. What could make that happen? R5 & R6 are the right values. Q2 seems to be working. Measure the voltage across R6: black meter probe on the end closest to IC1, red meter probe on the other. If the circuit was working right, we'd see about 6.6V. Another possibility: C2 might be shorted.

Geno,
The collector current is Q2 is (Vcc-Vref) / R5 = 33uA. HFE is in the 100's when Ic = 2mA. HFE decreases at low Ic, but let's just assume it's 100. That means the base current is 330nA. The voltage drop in a 1M resistor is 330mV. The voltage drop in a 6.8M resistor would be 2.24V. That would cause the emitter voltage to be about 2V lower compared to a 1M resistor and consequently, IC1 pin 7 would also be about 2V lower. Either way, the circuit works fine.
Click to expand...


I couldn’t seem to get any voltage reading on R6 on the bottom pcb pads or the top resistor legs.
Happened to have another 220k here so I swapped it. Nope. Swapped C2, nothing.

Reading about 3.4MV max if anything Pads of both inspected under magnification and photographed. Pics are not so great but trust me all looks just fine.
 

Attachments

  • 0AD33C4D-9F7F-45FD-8F95-4ADB7E4FB822.jpeg
    0AD33C4D-9F7F-45FD-8F95-4ADB7E4FB822.jpeg
    315.8 KB · Views: 17
  • B2800C88-6379-43CA-855C-2EBF50D1DDC1.jpeg
    B2800C88-6379-43CA-855C-2EBF50D1DDC1.jpeg
    243 KB · Views: 17
Last edited:
Upvote 0 Downvote
A few mV makes sense given the anomalous readings at the IC pins. Swapping parts is going to wear out the board. Let's just make measurements for now. Wait until we know a part is bad before you replace it. Clean the board, look for debris. Then with the power off, measure the resistance across C2. Should be a large number, but might be shorted somewhere. Then measure resistance from Q2-E to IC2-7. Measure & report both polarities. Should be 220K, might read less due to IC1. If it reads over 220K, then we have a bad resistor or a broken trace. Try to look under IC1, see if there is debris.
 
Upvote 0 Downvote
Moltenmetalburn said:
I couldn’t seem to get any voltage reading on R6 on the bottom pcb pads or the top resistor legs.
Happened to have another 220k here so I swapped it. Nope. Swapped C2, nothing.

Reading about 3.4MV max if anything Pads of both inspected under magnification and photographed. Pics are not so great but trust me all looks just fine.
Click to expand...
You really need to show Current full size Pics of your Build, These don't show on your earlier Pics & the one in the Blue Circle is not good.
The ones in red are also questionable:

Debugging.jpg
 
Upvote 0 Downvote
music6000 said:
You really need to show Current full size Pics of your Build, These don't show on your earlier Pics & the one in the Blue Circle is not good.
The ones in red are also questionable:

View attachment 4420
Click to expand...

those were still populated in the earlier pics.

I can’t as of yet the forum forces me to resize them. This is “large”. I need to figure out better pics for sure.

What you see are the empty pads from the component I removed. The pic was supposed to show the pads and traces are fully intact in those areas.

I can find nothing under magnification that look suspect. Been looking at this for HOURS. The pads have all been reflowed more than three times.
 
Upvote 0 Downvote
Chuck D. Bones said:
A few mV makes sense given the anomalous readings at the IC pins. Swapping parts is going to wear out the board. Let's just make measurements for now. Wait until we know a part is bad before you replace it. Clean the board, look for debris. Then with the power off, measure the resistance across C2. Should be a large number, but might be shorted somewhere. Then measure resistance from Q2-E to IC2-7. Measure & report both polarities. Should be 220K, might read less due to IC1. If it reads over 220K, then we have a bad resistor or a broken trace. Try to look under IC1, see if there is debris.
Click to expand...

At C2 my meter slowly climbs to 2000K and then stops/flashes back to zero when it reaches what I think is my meters max limit.

I have “IC1” as the tlo71and “IC100” as the charge pump chip. I measured the tlo71 pin 7. ??‍♂️

black at opamp end was doing same thing as C2.
red at opamp was 1487K.

I pulled opamp and inspected with bright light under magnification the best I could. I see nothing that looks less than perfect from any angle. No debris, no flux, looks pristine. Milled socket well seated.

I guess this means bad resistor/broken trace as you said earlier?
 
Upvote 0 Downvote
Upvote 0 Downvote
Moltenmetalburn said:
PRP 9372. $$$.

Precision Metal Film Resistors | Precision Resistive Products Inc | USA | Mediapolis

Precision Resistive Products Inc Axial leaded, Surface mounted, Thin Film, Thick film, Audio resistors, Current Sense, Surge protection, Melf, Pulse withstanding, High voltage, Array's, Networks, Matched sets, Microwave, SIPs, Power resistors, Custom value, AEC-Q200 automotive grade resistors
www.prpinc.com www.prpinc.com
Click to expand...
interesting.. they look nice. i assume that's mostly for aesthetics?
 
Upvote 0 Downvote
Upvote 0 Downvote
Moltenmetalburn said:
Q2-E to IC1-7 seems to do the same thing as C2, climbs to 200K then flashes to 1.

Why do I need to swap the multimeter leads to get consecutive readings?
Click to expand...

Because you're measuring the resistor while it's still connected to everything else in the circuit. The idea is to remove as much uncertainty as possible.

You got it fixed yet? The discussion seems to have gone off the rails, so maybe we're done?
 
Upvote 0 Downvote
Chuck D. Bones said:
Because you're measuring the resistor while it's still connected to everything else in the circuit. The idea is to remove as much uncertainty as possible.

You got it fixed yet? The discussion seems to have gone off the rails, so maybe we're done?
Click to expand...

no, not at all, I just didnt look at it for a few days while distracted by some other things and then picked it back up to get that last “measurement” yesterday.

You last asked: “...Measure resistance from Q2-E to IC1-7 with power off. Since IC1 is in a socket, make the measurement with IC1 out.”

I did that and said:

“Q2-E to IC1-7 seems to do the same thing as C2, climbs to 200K then flashes to 1.”
 
Upvote 0 Downvote
You must log in or register to reply here.

Similar threads

Quirkey
  • Question
Super Heterodyne: Noise but not the good kind
Replies
1
Views
212
jcpst
J
C
  • Question
Blue Breaker help - weak signal
Replies
8
Views
763
steviejr92
steviejr92
Z
Heavens to Murgatroyd! First 10 builds after a 20+ year hiatus. Mayonnaise, Devi Ever, etc, etc
Replies
6
Views
555
Feral Feline
Feral Feline
J
TUTORIAL Why Your Clean Blends Suck ... or "How to (Clean) Blend"
Replies
8
Views
1K
jrhevron
J
Svenson007
  • Question
Duocast testing help!!
Replies
3
Views
806
Mcknib
Mcknib
Back
Top