Bearhug Debug

[Kipper4]

Hi yall
last night i made Jon Patons bear hug
and it's behaving oddly
EG when i power it up it make a very loud buzzing noise which quiet quickly goes.
The its loses all volume within about 30 seconds.
Its almost like as soon as one of the caps get charged up it stops working.
substitutions
Q3        No MPSA18 in stock so I used an 2N5089 as suggested by Mr Paton himself.
For D3 and 4 i used 1n60's

R4 used a 4k7 as suggested again by Jon.

didnt have any 1k reverse log pots so i used a 2k pot with a 2k2 resistor across lug1 to 3. Its all i could think of.

Heres the thread
I used the perfboard layout

http://www.diystompboxes.com/smfforum/index.php?topic=99892.0


voltages
battery 9.2v    (after si Diode D1 voltage is 8.6v)
Vb 5.66


Q1
D 3.11
G 0.54
S 3.11

Q2
C 7.74
B 0.63
E 0.39

Q3
D 0
G 0.63
S 0

Heres some links to my dropbox pics


https://www.dropbox.com/sh/7ma9n53c5jqdgsr/d-0TvwYXDL#lh:null-Bearhug%20back.jpg

https://www.dropbox.com/sh/7ma9n53c5jqdgsr/d-0TvwYXDL#lh:null-Bearhug%20front1.jpg

https://www.dropbox.com/sh/7ma9n53c5jqdgsr/d-0TvwYXDL#lh:null-Bearhug%20front%202.jpg

i measured the DC voltage on input and output too just incase input 0v  Output 3.26v  is this normal or expected?

[duck_arse]

I think gate q1 should be about half v+, but I don't know what effect yr meter will have with that 10M gate resistor there. check your voltage at r5/r6, definatelty should be half V+ there.

wot value is that big green cap, where the 4u7 goes? wot voltage rating to make it so big?

I hate dropbox, and it's time for bed.

[Kipper4]

green cap is 4.7uf 100v
V bias is 5.66v at junction of R5 R6
sorry about the drop box i've yet to come to terms with embedding images
Night Duck Arse and thanks

[midwayfair]

1) I'm trying to figure out what's going on in the upper right of the back of the board. What is the tube jumpering over? There aren't any jumpers on the layout. Was your 10uF cap soldered into the row where the ground line was passing through?

2)

EG when i power it up it make a very loud buzzing noise which quiet quickly goes. The its loses all volume within about 30 seconds.

The behavior you're describing is what happens when the MOSFET fails to bias. Since your drain voltage is too low (by about 2-2.5V), and your source voltage is too high (by about a volt), that makes complete sense. This is also causing your gave voltage to be too low.

The good news is that you used sockets, which makes troubleshooting easier.

First: Pull out Q2. This will isolate the circuit from the rectifier.

Second: to eliminate the possibility that the FET is bad, you can replace it with a 470R resistor (or 1K, or even a piece of wire).

If you want to check that there's nothing wrong with your BS170, you can put a bipolar transistor in Q1. Go ahead and use one of the 5089s you have lying around (remember the pinout will be reverse). You should get LOUD and probably distorted signal. A high-gain transistor should bias somewhere at least close to the 5v range with the values used.

If you don't get loud signal, it's time to turn on the continuity setting of your multimeter and start checking all your contact points. Once you're satisfied that a part is connected to everything it's supposed to be connected to, then you start making sure it's NOT connected to something it SHOULDN'T be connected to (so just run the probe over all the contact points in the vicinity). While you're at it triple check all your resistor values.

If you DO get loud signal, double check the BS170. Maybe even flip it around in case for some reason yours have a reverse pinout. Use another one if you have it. Sometimes they blow up if they weren't in a static-proof baggie.

[Kipper4]

ive double checked the resistor values and all seem fine according to the chart calculator i consulted online
i removed Q2 and replace the Q1 with a bjt 5089 and it didnt lose volume or fade
i tried a couple of differant BS170 and the same old thing volume fade bad biasing etc
i removed Q1 and put the DMM on the sockets
D 8.64
G 0.83
S 0


I put the BS170 back in and put the meter probes on the Drain as i powered it up the voltages where nearer to what you said
D 4.5 ish
G 0.6
S 2.1

but after a few seconds the drains voltage falls back to 3.11v
and the sources voltage rises to 3.11

Stumped now??

and the plastic sheathing is over the grd wire because like you said i put the cap in the path of it.
there is no jumper as such its just protectioni even removed it to see it it was the problem and same thing

[bluebunny]

I'm not sure if this has any bearing on anything at all, but I think there may be a typo in Jon's perf layout?  If you look in the area of Q3 and read the three resistors to the right, they go 1M, 1K, 1K - which is what you have, Rich.  But I think it should be 1M, 1M, 1K?  (According to the schematic, there should be two 1Ms connected to the gate of Q3.)  What do you think, Jon?

Edit: You might want to ignore all that...      I just re-read through the whole thread and it looks like the original perf layout on Jon's original post now reflects the Version B and/or C component changes.  Long day.  Time for bed!   

[Kipper4]

yep i see what you mean.
I'll await Jons response before changing anything but i'll double check the old BOM too from Josh
Thanks for chiming in. All contributions are welcome.

[bluebunny]

Check my edit, Rich - posted simultaneously with yours.  I'm talking b0ll0x again...  

Edit:  If it helps, I have a vero layout that works!

[Kipper4]

I haven't had too much luck with vero but I'd sure like to try this compressor out for myself so yes please Bunny vero would be a welcome build for tomorrow
Cheers
Rich

[midwayfair]

I put the BS170 back in and put the meter probes on the Drain as i powered it up the voltages where nearer to what you said
D 4.5 ish
G 0.6
S 2.1

but after a few seconds the drains voltage falls back to 3.11v
and the sources voltage rises to 3.11

Stumped now??

Yeah, that's bizarre and I've never encountered anything like it.

Anyway, if it works with a bjt, then the connections are right. I didn't see anything wrong in any case.

Try pulling q3, taking voltages again, then replacing it with a resistor or jumper. See if the voltages stabilize.

I'm wondering if there's a possibility one of the electrolytic caps is leaking.

[jimilee]

A 2k pot with a 2k2 resistor? Can you explain to a noob how that works?

[Kipper4]

Best go read Geofx life of pots Jimilee.
In order to make a 1K pot out of a 2K pot i soldered a 2k2 resistor across from lug1 to lug 3 effectively halfving the pot value to a now measured 1k2 pot. I dont think it would magically turn it into a reverse log taper but the value was nearer to that desired by the parts list.
I hope this helps Jimilee

[duck_arse]

@ jon (is that the correct "attention, jon, question to you" syntax?)

on a working unit, what is the voltage at the gate end of the Q1 10M with and without the mosfet fitted? is this reading affected by meter resistance?

[midwayfair]

@ jon (is that the correct "attention, jon, question to you" syntax?)

on a working unit, what is the voltage at the gate end of the Q1 10M with and without the mosfet fitted? is this reading affected by meter resistance?

With:
D 5v
G 2.5v
S 1.8v

Without:
D 9v
G 2.7v
S 0v

[bluebunny]

If it's any use, Rich, here's the vero:

(clickable)

Edit: This is using Jon's original "V1" circuit.

[Kipper4]

my voltage are nothing like that theyre as reported earlier.
Any idea which cap might be causeing a problem if any Jon?
My guess would be one of the ones attached to the Q1 either the 10uf above Q1 on the drain or the 47uf attached to the source of Q1
thanks for any hlep.
If i cant debug this i'll give the vero a whirl Mark thanks.

[midwayfair]

Have you check continuity on all the components associated with q1? There could be a bridge or something we can't see. You really can't trust your Reyes all the time.

[Kipper4]

I'm jiggered if i know Jon
i checked continuaty from everything attached to the mosfet and replaced a 47u cap and 2k7 resistor and its still behaving oddly.
loud hum on power up. works for a start then volume tapers off to hardly there at all and it sputters on its way out.
if i put my finger under the pcb to activate the mosfet the circuit works
it's driving me crackers.
Thanks for all the help.

[midwayfair]

if i put my finger under the pcb to activate the mosfet the circuit works
it's driving me crackers.
Thanks for all the help.

Check ALL your ground connections. Make sure the 2.7k is grounded especially. Reflow solder to stuff connected to the source.

[Kipper4]

I did some more checking and i cant figure it out yet i will. I'm having a day off from debugging the perf. Meanwhile i built the vero and its not exactly tonaly transparent infact i'd say its a lot more bassy.
maybe later when i get s=done at work i'll post some pics of it too
thanks