Amps Hot Fender Pro Jr

CapnDenny1

Student Of The Blues
OK, time to do something else for a bit. But here is what I found out today. It is still a mystery, but I am not crazy.

I put a dmm on the hv secondary and set it to AC volts. When I powered it up it was rock solid at 250vac.

I put a second DMM on the B+, and while I watched the AC hv stay constant I could see the B+ steadily rise.

I put a scope probe referenced to chassis ground on the point where the hv secondary connects to the diodes. One channel on each of the leads. I could get a similar result by adding a DC offset into the output of the transformer output in my pspice simulation. I captured it after a few seconds and then waited a few more seconds and halted the scope and took a picture. You can see the earlier voltage levels on screen in white.

IMG_2806 by Dennis Kelley, on Flickr

The scope trace in yellow was the one connected to the same side as the bias circuit coupling cap.

Here is the schematic of my pspice sim.

bias sch by Dennis Kelley, on Flickr

And what the scope traces are supposed to look like.

bias good sim by Dennis Kelley, on Flickr

I reasoned that what appears to be happening is the AC hv output of the transformer, while staying at 250 vac between them, is getting a DC offset. So I added a cap on the bias ckt side, and it kind of does it. Not quite. But notice here the pspice output looks more like my scope shots.

bias sim by Dennis Kelley, on Flickr

Of course that cap does not exist in the circuit, or at least it isn't supposed to. I am thinking perhaps the power transformer has some sort of issue?

I put the variac on it, thinking perhaps it only happens at high voltage. But not, it drifts the same way with the AC line at 60 vac. It does stop eventually, at low voltage and full voltage.

So I am going to go stew for a bit, maybe play guitar or watch a little TV. My brain is tired of trying to understand things that don't make sense. You would think I would be used to it by now? But I just can't stand not knowing why.
 

OG_Blues

Guitar Geezer
It's not often I have almost no idea what might be causing a malfunction, but this is a very odd one.
I suspect you may be on to something suspecting the power transformer, but I can't imagine what in a transformer could go wrong to cause that. I have never observed anything like that from a transformer - there's always a first time.
Since you have a steadily but slowly rising B+ voltage, that tends to indicates that there is a rising voltage somewhere earlier in the power supply circuit.
Find the first "series node" in the power supply where the voltage is rising, and you likely have the answer.
I would start at the transformer primary, then the transformer secondary, then systematically at each successive node in the circuit to identify the first 'source" where the voltage behaves this way.
Since you have varying voltages with both the B+ and the bias, this tends to point towards the transformer also.
Also, since you seem to have a distorted waveform, use the scope to identify the first point in the circuit where you observe an abnormal waveform.
It will be interesting to see what you eventually determine on this one.
 

CapnDenny1

Student Of The Blues
The thing that is screwy is it behaves normally if I disconnect the bias voltage cap.

The transformer hv output line to line stays constant at 250 vac, regardless.

How does the transformer know that the bias circuit is there?

Somehow that bias circuit, when operating, shifts the dc offset of the hv secondary.

First thing I will check is to see if this is just an artifact that is there because there is basically no load onth hv during most of this. With tubes all pulled there is nothing tokeep the voltages honest. It one thing to cause a circuit that is floating to drift, and another to be able to drive a load. I think it still hppens with tubes. But since I can see it at lower voltage I can use the variac to test at 60 volts with the tubes in, and not isk damaging them.
 

CapnDenny1

Student Of The Blues
Some progress. I was experimenting with the circuit in the simulation trying to duplicate the results. By thinking it through I got an idea that I wondered what would happen if one of the diodes didn't work very well. So I tried leaving this one out, then that one, etc. After about an hour of frustration I put a 50 DC power supply in series with one of the diodes. The one pointing to the side that feeds the bias circuit. I got this result from the simulation.

bias sim by Dennis Kelley, on Flickr

As compared to the actual scope capture of the same points.

IMG_2806 by Dennis Kelley, on Flickr

Looks pretty close to me.

I tested the diodes in circuit with the diode checker part of the DMM and all four tested out OK. But the circuit starts out working OK as well. I am thinking the bad diode has too much resistance, or is breaking down, or something. It doesn't explain the bias circuit tie in. Why it only drifts when the bias circuit is connected in?

The schematic calls for 1n5062 diodes, but it actually has 1n4006 diodes. I ordered some 1n4007 which have 1kV breakdown, slightly higher than 1n4006. They are guaranteed by Wednesday. So I am hopeful this is it. Maybe I will find a broken trace when I replace them, or a short. This amp has been butchered up pretty bad.

It's kind of a scary thought that a failed diode, a $0.05 part, has the ability to raise the intended B+ from 320V to 500V and simultaneously reducing the bias voltage from -11V to -4V. You could not create a better way to fry a power tube than increasing the voltage, and increasing the current through it. I guess you could hit it with a hammer maybe?

Stayed tuned.....
 

CapnDenny1

Student Of The Blues
I am thinking there is a flaw in the design of the Fender Pro Jr amp. The hv secondary is floating and they connect it to a full wave bridge circuit. The output of the bridge provides the B+, with help from the filter caps. That works just fine, as long as I don't connect the power tube bias voltage circuit to the hv from the transformer.

There is nothing special about the bias voltage. Typical Fender capacitvely coupled with a diode to provide the negative bias voltage, and a couple of caps to filter it, and resistors to set the voltage to -10.5V.

When I first connect it to the hv winding of the secondary, I see the -10.5V and B+ is around 350V, which is fine according to the schematic I have. However after a few seconds it starts to drift. The B+ starts to climb, and the bias voltage goes down towards 0V, down around -4V. The B+ goes up to 500V.

I have replaced or tested every part in the bias circuit. I experimented with what it takes to get it to start drifting. The coupling cap goes to a 58K resistor which is tied to ground. If I disconnect that or the coupling cap, then the B+ stops drifting.

What appears to be happening is that the hv winding on the power transformer starts to develop a DC offset. Not between the hv output leads line to line, but between them and the GND reference. A common mode voltage is somehow getting imposed onto both of the hv windings. You can see the scope capture waveform at the hv output leads of the PT above.

Somehow this offset gets added to the B+, and only 1 set of diodes in the bridge ever get turned on. The other diodes never conduct because the voltage on the positive cycle never reach the B+ voltage so they aren't forward biased. When that side goes negative it doesn't make it to GND or even try to make it below GND so the other diode doesn't turn on either. The diodes are all good mind you. They tested good, but I replaced them anyway.

This looks like a design flaw in the amp. I have ordered another PT out of frustration and a lack of what else to try. Hopefully that will fix it, but I am waiting for the transformer.

I reasoned that the bias circuit is causing some sort of imbalance, and the coupling cap is getting charged up. I think this has the effect of pushing the common mode voltage of the PT hv outputs. So I built a quick little circuit with the same 0.047uF cap and a 56K resistor to GND. I connected it to the hv output on the other side of the PT secondary. To my surprise this actually worked. So with a duplicate circuit pushing it back on both sides the DC offset stays near zero. The B+ goes to 320V and the bias voltage goes to -10.5 volts, and everything is stable.

Even though this works, I shouldn't have to do this. Fender knows what they are doing right? I'm not sure if I've ever seen a secondary B+ circuit that had no DC tie to GND at all. Usually they have a secondary with a center tap that is grounded. This has nothing. And as I said it works fine, until you hook up the bias voltage circuit.

Let's hope the PT fixes it. I would prefer not to have extra circuits added to a customer's amp.
 

david moon

Attempting the Blues
I am thinking there is a flaw in the design of the Fender Pro Jr amp. The hv secondary is floating and they connect it to a full wave bridge circuit. The output of the bridge provides the B+, with help from the filter caps. That works just fine, as long as I don't connect the power tube bias voltage circuit to the hv from the transformer.

There is nothing special about the bias voltage. Typical Fender capacitvely coupled with a diode to provide the negative bias voltage, and a couple of caps to filter it, and resistors to set the voltage to -10.5V.

When I first connect it to the hv winding of the secondary, I see the -10.5V and B+ is around 350V, which is fine according to the schematic I have. However after a few seconds it starts to drift. The B+ starts to climb, and the bias voltage goes down towards 0V, down around -4V. The B+ goes up to 500V.

I have replaced or tested every part in the bias circuit. I experimented with what it takes to get it to start drifting. The coupling cap goes to a 58K resistor which is tied to ground. If I disconnect that or the coupling cap, then the B+ stops drifting.

What appears to be happening is that the hv winding on the power transformer starts to develop a DC offset. Not between the hv output leads line to line, but between them and the GND reference. A common mode voltage is somehow getting imposed onto both of the hv windings. You can see the scope capture waveform at the hv output leads of the PT above.

Somehow this offset gets added to the B+, and only 1 set of diodes in the bridge ever get turned on. The other diodes never conduct because the voltage on the positive cycle never reach the B+ voltage so they aren't forward biased. When that side goes negative it doesn't make it to GND or even try to make it below GND so the other diode doesn't turn on either. The diodes are all good mind you. They tested good, but I replaced them anyway.

This looks like a design flaw in the amp. I have ordered another PT out of frustration and a lack of what else to try. Hopefully that will fix it, but I am waiting for the transformer.

I reasoned that the bias circuit is causing some sort of imbalance, and the coupling cap is getting charged up. I think this has the effect of pushing the common mode voltage of the PT hv outputs. So I built a quick little circuit with the same 0.047uF cap and a 56K resistor to GND. I connected it to the hv output on the other side of the PT secondary. To my surprise this actually worked. So with a duplicate circuit pushing it back on both sides the DC offset stays near zero. The B+ goes to 320V and the bias voltage goes to -10.5 volts, and everything is stable.

Even though this works, I shouldn't have to do this. Fender knows what they are doing right? I'm not sure if I've ever seen a secondary B+ circuit that had no DC tie to GND at all. Usually they have a secondary with a center tap that is grounded. This has nothing. And as I said it works fine, until you hook up the bias voltage circuit.

Let's hope the PT fixes it. I would prefer not to have extra circuits added to a customer's amp.
I was going to suggest excessive (and changing) reverse current in the diode, but you replaced that. In your "actual" scope traces, I don't understand the flat top on the secondary output. Also if I am understanding correctly the green trace is the B+ which looks nothing like the Spice, which is DC with some ripple (what I would expect).
 

CapnDenny1

Student Of The Blues
The scope pics are the output high voltage wires from the power transformer.

The flat top and flat bottom are where the diodes are tuned on. The tubes were removed so there isn’t any ripple.

I spent a couple more hours today trying to understand what is actually happening. I finally just set it aside while I wait for thePT to arrive.

I have 5 other amps, or is it 6? Anyway, it’s not fair to make everybody wait, just because I’m too dumb to fix one.
 

DannyB

2 miles from Jim Beam. Oh! Pleasent Hope!
Hmmmmm...intrigue indeed!

1volt, per 60 seconds on a 60hz transformer.

I'm going to have to buy you two beers before this is over.
 

CapnDenny1

Student Of The Blues
I got the replacement PT in andno difference. I also tried adding the 0.01uF caps across the rectifier diodes that is in the newerPro Jr’s, also no help.

So out of a lack of other ideas I added a cap and resistor from the other hv secondary wire to gnd. That way both hv leads ee the same load. I put heat shrink on it and hot glued the cap to the filter caps, so it’s stable.

With that fix the B+ is rock solid right at 325v as it says on the schematic. The bias is stable as well. I texted the customer and told him he owes me for 4 hours work. I replaced the sockets on the power tubes, but mostly it was trying fix the runaway voltage.

I played it thru my Carvin Nomad’s speaker, which I use as a test speaker when the amps are on the bench, and it sounded pretty nice. Once I put it back in it’s own cab, the 10” Celestion in the Pro Jr didn’t sound as good. I will recomend to him to get a better speaker.

I also fixed that darn Legend A30. Now I can get to some of the other amps. I have 6 that have been waiting on these two.
 

CapnDenny1

Student Of The Blues
I called him back and told him I wanted to figure out what was going on. I explained that I had other customers I had to take care of, but I wanted to get to the bottom of his amp’s issue. He agreed.

I have a spare PT so I am going to mock up the circuit on a breadoard and see if has the same issue.

I will also take more parts off the board and se if I have some pc board issue.
 

david moon

Attempting the Blues
If there is a PC board involved it could crack and then your schematic may not reflect what paths are actually there (or not). I encountered that once with a Roland Jazz Chorus
 

CapnDenny1

Student Of The Blues
Ya, I had that too. But so far I haven’t foud anything. It’s a simple board, and should not be this hard.
 

CapnDenny1

Student Of The Blues
Mystery solved. Once again I have failed at trying to fix an amp that isn’t broken!

Using Pspice I was able to simulate the issue, with the transformer secondary getting a dc offset. But in my simulation it only did it when there was almost no load on the transformer. Once I added even as much as 1 preamp tube the problem did not happen.

I was pretty sure I tested that and still saw the overvoltage, but I wanted to give it one more check. So I removed my extra circuit and monitored the B+ and flipped on the power, with all tubes present.

It quickly went to 300, 320, 350, then 345, 340, 330,and settled in right about 320, right where it needed to be. The bias voltage was also right at -10.5v where it is supposed to be at.

I think when I tested it, I saw it start to climb and shut it off right away. So the reason for the runaway tubes was the bad tube sockts all along as I originally determined. That danged Legend 30 that I burned up the PT and cost me $175 just has me gun shy!
 
Top