|
Vintage Audio (record players, hi-fi etc) Amplifiers, speakers, gramophones and other audio equipment. |
|
Thread Tools |
19th Sep 2017, 7:25 pm | #1 |
Rest in Peace
Join Date: Jan 2015
Location: Cottesmore, East Midlands, UK.
Posts: 858
|
Fender Hotrod Deluxe Parasitic Oscillation
Hi, All,
Had a friend's Fender HRDX (40W Push-Pull guitar amp) in for a servicing. These amps are prone to a number of stock faults. Having addressed those and modified the Master Volume control (basically replace the stock linear pot with a logarithmic pot), I connected up an audio function gen and the scope to give it a workout. With the ear defenders on and the wife out walking the dog, I fired it up and all seemed well. I was running on the "Clean" channel and monitoring the scope as I wound the volume control up. Looked good until I got to about 8 on the volume (this particular Fender model goes up to 12, one better than Spinal Tap!), when a large wart appeared on the trace, shown in the first picture below. The upper trace is monitoring the output of the preamp, the lower trace is across the loudspeaker and shows a large burst of parasitic oscillation round about the maximum negative portion of the waveform. Since the output of the preamp is still clean, it looks as if the problem lies in either the phase splitter or the output stage. The second image shows the output stages of the HRDX. I moved the scope leads to the junctions between C26/R64 and C27/R65 ie between the output stage coupling caps and the grid stoppers on the OP valves - both traces were clean, thereby eliminating the phase splitter. At this point, despite the ear defenders, my ears were starting to ring and small objects were walking off the workbench, so I shut down and replaced the speaker with an 8R 100W dummy load resistor. I repeated the above checks in a bit more comfort and blow me if the problem hadn't disappeared. The trace across the dummy load was as clean as a whistle up to 10 on the volume control when the normal clipping started to appear, but no trace of the parasitic oscillation at all. I shut down, reconnected the speaker, fired up - and there was the problem, right back again! Playing around with it, the oscillation is most pronounced on lower frequencies - the image below was taken at 150Hz. On increasing the input frequency the onset of the oscillation shifts upward, so that at 1KHz, you need a volume setting of 10 before it starts; and when it does it is not so pronounced. At this point I tried replacing the output valves - no difference. Any theories/suggestions? Cheers, Frank Last edited by frankmcvey; 19th Sep 2017 at 7:38 pm. |
19th Sep 2017, 8:29 pm | #2 |
Dekatron
Join Date: Feb 2007
Location: Leominster, Herefordshire, UK.
Posts: 16,535
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
If it's clean on a resistive load you could try a Zobel network across the speaker.
Measure the speaker L&R and work out where 6.28*f*L =R. Get another watty R of same value and connect in series with a capacitor whose reactance at f is also the same as R. The parallel combination of speaker and Zobel should look resistive all the way up give or take. Connect across speaker or amp output and see...... Alternatively you could try assorted grid and anode stoppers. Does the amp have any NFB? If so, is it stable open loop? If it is then playing with the feedback may help (though if you're not familiar with loop stability criteria it will probably make things worse!).
__________________
....__________ ....|____||__|__\_____ .=.| _---\__|__|_---_|. .........O..Chris....O |
20th Sep 2017, 9:26 am | #3 |
Nonode
Join Date: Oct 2008
Location: Warsaw, Poland and Cambridge, UK
Posts: 2,677
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
Looking at the circuit diagram, there is a feedback path from the 4 ohm tap to a network involving the 'presence' control. When you can see the parasitic oscillation, does twiddling the presence control make any difference?
Chris
__________________
What's going on in the workshop? http://martin-jones.com/ |
20th Sep 2017, 8:44 pm | #4 |
Rest in Peace
Join Date: Jan 2015
Location: Cottesmore, East Midlands, UK.
Posts: 858
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
Hi Chris & Chris,
Thanks for the suggestions. Chris Herald. Speaker - R=7.5 Ohms, L= 0.4mH. Doing the number crunching, I came up with a frequency of around 2985 Hz. Feeding that into the Xc equation gave me a capacitance of 7.1uF. Assuming I've got my sums right, the only caps I have near to that value are all electrolytics. I think it would probably be quite a lossy solution as well, and, while my mate is more likely to be playing The Dog & Duck rather than Madison Square Gardens, I don't suppose he'll want to lose too much headroom! Chris Jones. Good thinking, that man. Yes, it does indeed. The "presence" control gives higher frequencies a boost by cutting down higher frequencies in the NFB, and it has a greater effect on the distortion channel than it does on the clean, since the distorted waveform is richer in harmonics. By turning the Presence control up to 4 (was originally at 0) the parasitic oscillation disappears altogether. The presence acts more like a conventional treble control when the clean channel is selected, and the extra treble generated by turning it up to 4 could be compensated by turning down the treble control in the preamp. More of a workround than a fix, but we'll give it a go. Cheers, Frank |
20th Sep 2017, 9:43 pm | #5 |
Nonode
Join Date: Jul 2010
Location: Folkestone, Kent, UK.
Posts: 2,172
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
Hi Frank
the grey ribbon cables are a known cause of instability, try dressing them by bending them in to a different position. Some folk have gone as far as splitting the cables or simply rewiring them, if that is the cause |
21st Sep 2017, 1:13 pm | #6 |
Octode
Join Date: Jan 2012
Location: Wimbledon, London, UK.
Posts: 1,465
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
I have cogitated on this for a day or two and I too thought that the NFB/presence control might be the root cause (but anyone could say that, given that other folk have already suggested it). You could try just disconnecting the wire going from the output transformer to the NFB point. If that wire is part of a multi-wire ribbon cable, then that could well be a cause of parasitic oscillation. If disconnecting it stops the oscillation completely, then re-routing it, or possibly splitting it out of the ribbon, could be a good idea. In my opinion, sending NFB via a ribbon-cable is a bad idea and a recipe for parasitic oscillation.
Another thought that I had was to try another speaker, although it seems unlikely that a working speaker should cause parasitic oscillation if there is no device to pick up the signal and feed it back into the amp somewhere, doesn't it? Good Luck, Colin. Last edited by ColinTheAmpMan1; 21st Sep 2017 at 1:15 pm. Reason: Correction of poor grammar. |
21st Sep 2017, 9:20 pm | #7 |
Rest in Peace
Join Date: Jan 2015
Location: Cottesmore, East Midlands, UK.
Posts: 858
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
Hi, ITAM & Colin,
Thanks for the replies - all good thought-provoking stuff! ITAM - I've done a bit if reading around the ribbon cable problem and I've taken your advice. Pic 1 shows the standard factory dressing of the ribbon cables; pic 2 shows the dressing method recommended by the Fender internet gurus. Note that the amplifier back has been removed - when it is fitted, the upper part of the ribbon runs - horizontal in this view - will lie against the foil shielding of the back. Unfortunately the fault is still apparent after the change. The NFB line is not routed through the ribbon cables; it is the grey wire coming from the external speaker jack, (left jack in pic 3) and heading off under the valve base PCB, which it joins at the point beside the pen tip in Pic 4. Apart from the ferret's nest around the jack itself, I don't think it would be possible to route the NFB any better, running as it does along the angle of the metal chassis as far from the circuit board as possible for most of its run. Colin - On disconnecting the NFB line from the jack and carrying out the tests again (at an ear-bleeding volume!!) the parasitic oscillation had disappeared, right up to 12 on the volume control. So it looks as if the NFB line is the culprit. On tidying the wiring at the jack - Pic 5 - and testing again, the oscillation was back again, right on cue. I don't see how I can improve on the routing of the NFB cable as it now stands. I've considered ferrite beads, but it looks from my first picture in the OP that the oscillation is actually at a high audio frequency rather than a radio frequency, so I don't think beads would help. Having said that, when cranking the volume up through the onset of the parasitic oscillation, there is only a very minimal change in the sound from the speaker. Be interesting to see how the sound varies with an actual instrument connected. I think my next step will be to replace the NFB grey wire with a length of miniature shielded cable and see how that goes. Cheers, Frank Note - Just because I happened to have a spare 15" 8 Ohm speaker lying under the bench, I tried that as well - exactly the same PO, I'm afraid.. |
21st Sep 2017, 9:55 pm | #8 |
Nonode
Join Date: Jul 2010
Location: Folkestone, Kent, UK.
Posts: 2,172
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
Hi Frank
the NFB tap goes to the phase splitter, try replacing the 12AX7 with a 12AT7? A bit radical I know, but it shouldn't affect the sound too much I dont think - well try it, baby and bathwater etc. |
21st Sep 2017, 11:44 pm | #9 |
Rest in Peace
Join Date: Feb 2004
Location: Hexham, Northumberland, UK.
Posts: 2,234
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
That 47pF capacitor between the phase splitter anodes (C30?) seems a bit of a suspicious add on, as if it may have been added as a fix for something during development. I would be tempted to increase the value a bit, just to see what effect it has on the problem. Also it may be worth trying decoupling supplies around that part of the circuit with something like a 0.1 microfarad capacitor, in case it is something to do with high frequency feedback through the power supply. All those diodes around the output stage seem a bit of an odd addition too. What are they all for?
Alan. |
22nd Sep 2017, 8:54 am | #10 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,866
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
There is a diode clamping each anode to stop it going negative. This doesn't sound like a problem in need of a fix, but through the action of the transformer (and neglecting leakage inductance effects) this will also clamp the opposite anode preventing it going above 2x HT voltage.
Guitar amps often connect to speakers via jack sockets. They short when someone fits or removes a plug, they go open when dirty or someone fires up the amp, hits a chord and realises they never plugged the speaker in! Valve amps can be damaged by voltage transients when driven into a disconnected load. To keep the universe fair, transistor amps don't get it any easier, they're b*****ed if driven into a short and need loads of protection. The one in the cathodes is there to stop the cathode resistor voltage rising and compensating for valve bias conditions a bit and thus risking a terrible increase in valve life. Oh and someon'll probably tell you it increases 'crunch' or something. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
22nd Sep 2017, 9:15 am | #11 |
Nonode
Join Date: Jun 2007
Location: Coulsdon, London, UK.
Posts: 2,163
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
The parasitic oscillations seem to be only affecting one half of the waveform.
Would there be any advantage in finding out whether V4 or V5 is associated with the oscillations? If you were to reduce the signal at one of the output valve grids by 'shorting' the signal to chassis with a 0.1uF capacitor you might be able to identify the troublesome side of the circuit. If either CR4 or CR5 were open circuit the amplifier would continue to work but the output circuit would be unbalanced. |
22nd Sep 2017, 9:51 am | #12 |
Dekatron
Join Date: Nov 2010
Location: Oxford, UK.
Posts: 4,980
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
Daft question - have you checked the valves?
|
22nd Sep 2017, 1:06 pm | #13 |
Octode
Join Date: Jan 2012
Location: Wimbledon, London, UK.
Posts: 1,465
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
Here's a silly thought....
If the amp has always had this problem, it could be that the NFB line has the wrong phase and is always positive. If that were the case, then reversing the wires from the output valve anodes to the output transformer might effect a cure. It certainly sounds (pun intended) as if the phase of the NFB is either always wrong, or just wrong at some setting of the volume control. Craig, I think Frank has already tried changing the output valves (post #1), but swapping the phase-splitter might be informative, I would agree. When Fender got taken over by CBS, there were some changes made to several of the amplifiers by the "new brooms". A lot of extra capacitors were put in to tame instability, but unfortunately much of this stuff also altered the "sound" of the amplifiers in a bad way. These amps were generally recognised as having silver control-panels, rather than black ("Silverface" and "Blackface"). Maybe your amp has a similar problem. You could also look at the routing of the wires between the phase-splitter anodes and the output-valve grids; a bit too close to one another and oscillation is almost inevitable. I would also suggest that the capacitor between the phase-splitter anodes is something to get rid of, once you solve the oscillation problem. At 47pF I would think it is only effective at RF, anyway, though I could be completely wrong on that. Colin. |
22nd Sep 2017, 1:15 pm | #14 | |
Octode
Join Date: Jan 2012
Location: Wimbledon, London, UK.
Posts: 1,465
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
Quote:
Colin. |
|
22nd Sep 2017, 1:18 pm | #15 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
I wonder if the anode diodes are chattering.
Lawrence. |
22nd Sep 2017, 2:38 pm | #16 |
Guest
Posts: n/a
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
I am just wondering about the anode diodes full stop, seems rather silly to me.
|
22nd Sep 2017, 3:37 pm | #17 |
Octode
Join Date: Jan 2012
Location: Wimbledon, London, UK.
Posts: 1,465
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
Not quite as silly as you might think, Merlin. The late Ken Fischer of Trainwreck amps suggested putting three 1N4007 diodes in series from the anodes of the output valves to ground (anode connecting to ground). This will suppress spikes at certain frequencies. Ken suggested that these particular diodes do not affect the sound of the amp, while other diode types do. I am not familiar with the particular diodes on the amp under consideration, but if they are fast recovery, I would replace them with the suggested 3 off 1N4007.
Another spike-suppression mod which helps to eliminate mains-borne spikes is to put a 250V M.O.V. across the mains transformer primary. Colin. |
22nd Sep 2017, 3:42 pm | #18 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
According to the schematic the diodes are R3000's
Lawrence. |
22nd Sep 2017, 4:08 pm | #19 |
Nonode
Join Date: Jun 2007
Location: Coulsdon, London, UK.
Posts: 2,163
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
If you get desperate you could try a series RC snubber network across the primary of the output transformer or across CR4 and CR5.
|
22nd Sep 2017, 4:46 pm | #20 |
Heptode
Join Date: Apr 2003
Location: Cornwall, UK.
Posts: 988
|
Re: Fender Hotrod Deluxe Parasitic Oscillation
I think I would try a resistor in series with the 'Presence' control - maybe 2K2?
SimonT.
__________________
The honesty of imperfection.......... |