Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 28, 2024, 12:01:54 PM

Login with username, password and session length

Recent Posts

 

Heatsink questions on Crate

Started by txflood, February 12, 2013, 12:33:41 PM

Previous topic - Next topic

txflood

Quote from: DrGonz78 on February 14, 2013, 04:49:04 AM
I would not replace those main filter caps on first check... Instead I would remove the solder and re-solder the existing caps to start. While I was in there I would solder up the Bridge or rectifier diodes, while also hitting solder joints on all the wire wound resistors to make sure those were solid. Also, hit the input jack solder joints as that might be the culprit right there to start. Let's see how many parts really need replacing and how many parts just need to be soldered w/ real leaded solder...

Edit: BTW what make model is this amp and what year do you think it was made?

I will give this a shot and see if it makes any difference.

It is a Crate RFX 65 that I think must have been made around 2005 from looking at the date codes on some of the components.

txflood

Well I think I've made some progress on this tonight. The ribbon cable that I was disconnecting goes to the FX control board and supplies power to it but no audio signal. That is J4 on the schematic.

When I unhooked J10 on the other hand, that stopped most of the annoying hum that I have been hearing. This connector has audio going to and from the effects DSP board along with DC power being sent to it that is rectified through diodes D31, D32, D33, and D34 on the main circuit board. The main filter cap for this DSP voltage supply is C64 1000uf.

The DSP board has lots of surface mount components and I don't have any schematics for it.

It looks like the output from the DSP board goes to pin 1 on J10 and from there is mixed with the selected preamp channel and CD input and then on to the power amp. Point D comes into play to have the signal inverted if there is a cable plugged into the input jack or it bypasses the inverting opamp IC4A if there is no cable. Either way it looks like the DSP output gets routed on to the power amp which would explain why I hear the noise even if there is no cable plugged in.

The wires running from J10 to the DSP board are not shielded which I'm sure is not helping but I'm not sure at this point if that really accounts for the noise. I'm stopping here for now.

DrGonz78

#17
Maybe my day was just crazy and my brain is mush.... Maybe that is everyday for me... Sounds like your referencing a schematic here and it would be nice to follow along in better detail. Can you send us copies through messages/email or post a copy on the thread? It will help greatly to follow along with you, thanks.

Edit: Sorry I thought was looking a schematic that might be the right one but was wrong. However, I think this was the right one?? Let us know if this schematic is the same as the one you are referencing...?

Double Helix Edit: Yeah that looks like the right to me... Look at the link to MEF website for the exact problem you described.
http://music-electronics-forum.com/t13806/
Read the last post for a possible fix.
"A person who never made a mistake never tried anything new." -Albert Einstein

phatt

Well spotted Dr G, :tu:

Yep my pet hate,  connector blocks designed for computer crap used in Audio circuits are a bad idea and asking for trouble.

I recall reading way way back many years a comment from a Teck chap who worked in radio stating if you want it to work reliably,,, SOLDER EVERYTHING.

My other favorite rule of thumb; 90% of all electrical problems will be found in connections.

Why just this week I revived a dead 22 inch computer monitor (about to go in land fill) by finding the dud connector which was hidden deep inside the harness.
One tiny bent pin in the connector to the Screen was causing the colour to drop out at random.
Works perfect now and I always wanted a big screen.  :dbtu:
Phil.

txflood

Hey it's been a while since I have been able to get back to this problem, but I wanted to provide everyone an update on where things stand. In my case I don't believe that I have a connector problem but rather a noise problem due to the unregulated voltage being sent down to the DSP board.

One of the things that I've done since I last posted was to turn an unused laptop into a primitive scope using the software available on the net plus a circuit buffer board. I wouldn't want to try to take precise measurements with this rig, but it at least offers some insights as to what is going on.

If you look at the schematic, there is a rectifier circuit built around 1000uf C64 that sends unregulated DC down to pin 4 of the connector on the DSP board. Once it arrives there it goes to a LM117 voltage regulator that outputs 3.3 V for the digital circuitry. This regulator has a typical 65dB rejection ratio if no adjustment capacitor is used. I can't really tell what components they have tied to the regulator and haven't tried to figure it out yet.

To show you the big picture, the red wires that are about 10 inches long run from the main amp board to the DSP board:



And here is picture of this multi-plane very dense SMD digital board:




So what I am seeing on my "scope" is about 1 Vpp 120 Hz ripple arriving on the DSP board DC input and going to the voltage regulator. This is producing a noticeable 120 Hz hum being being present all the time coming out of the amp because of the DSP output always being mixed in. This equates to a flat low-B humming out of the amp and what has been driving me crazy enough to try to do something about it.

One thing that I found that noticeably reduced the hum was to add a 4700uf capacitor across C64. This gets the ripple down into the .200-.300 Vpp range and seems to get the hum down to a tolerable range where just the broadband type of noise coming from the DSP board is more present.

This looks like it might be a good solution in replacing C64 but one thing I'm not sure about is the purpose of the other circuit that C64 feeds. If you look at the schematic to Q31 and Q16, I think what is going on there is that this is supposed to effectively ground the input to the power amp when the power is turned off to kill wacky artifact sounds as everything fades out. I may be totally off on that so I was hoping someone else here could comment on the intent of that circuit.

One other thing that I tried was to physically separate the red wires going to pins 1 and 2 on J10 from the wire going to pin 4 as much as I could. I was thinking that perhaps there is some significant noise coupling going on along the 10 inch run of unshielded red wires, but I can't say that this made a noticeable difference. I was thinking that running shielded wire with the shield grounded on one end for the audio signals might help here, but I'm not convinced at this point that it would be worth the trouble.

The other idea I was toying with is to add either a pot or a switch between J10 pin 1 and R92 to ground the DSP output (maybe jumper wire JOW1 is the ideal spot) so that when I just want to get the amp as quiet as possible and use external effects, I kill the DSP board output signal. This would still leave the capability to turn the internal DSP back on when the situation allows it.

Years ago I had some experiences dealing with RF noise and personal computer digital communication equipment and even though we had lots of theories on suppressing things, it always seemed to turn into a long tedious trial and error process to find the right combination of things to pass the interference tests. I know this is an audio circuit with much lower frequencies but I tend to think this could also waste a lot of time with diminishing returns trying to do too much with what is probably just a noisy DSP board to begin with.


Any thoughts?

Enzo

Maybe i am just dense, but I can't find what you are looking at.  Can you link the schematic or tell me exactly what drawing numbers you have?  You mentioned RFX65, but Gonz linked GFX65.  And I cannot fins half the parts you mention.  For example I see C64, but the DSP cards have 5v regulators.  And C64 only goes to teh DSP conector on this print.  I can find Q16 as a mute, but cannot see a Q31.  And Q16 doesnl;t appear to conect to C64.

Raw voltage like C64 is normal to have some ripple.  What matters is how much ripple is there after the regulator.  The DSP doesn;t run on the raw voltage. 

I think Roly touched on this.  It may help to over filter at C64, but I suspoect you have a ground issue, and the filtering ground return currents are getting onto the DSP signal.  If I recall, the DSP card has four corner poip-in standoffs.  Is there a grounding screw on any corner?

DrGonz78

#21
Yeah I think I might have screwed the pooch there...   xP
I thought the file I had loaded was right but just had the wrong named file....

I think this file is correct and maybe the OP can confirm this schematic is correct or post a copy?

Edit: Enzo is right on with these amps having a grounding screw on one of the DSP board standoffs. That needs to be looked at too.
"A person who never made a mistake never tried anything new." -Albert Einstein

txflood

Sorry I didn't look more closely at the previous schematic that was posted. Dr Gonz has posted the correct one now.

I believe on this amp there were 4 plastic standoffs in each corner of the DSP board so I don't remember any kind of chassis ground connection or screw. I will take another look this evening.

Enzo

Q16 Q31 are just a power amp mute.

Roly

Quote from: EnzoI think Roly touched on this.  It may help to over filter at C64, but I suspect you have a ground issue, and the filtering ground return currents are getting onto the DSP signal.

I did?  I don't recall; but I was a bit surprised that the recto is on the same board, and I certainly agree that it could be a grounding issue.  This digital stuff is much much more like RF than audio and grounding digital earths and ground planes can be much more critical.
If you say theory and practice don't agree you haven't applied enough theory.

txflood

Quote from: Enzo on March 26, 2013, 10:08:01 AM
Q16 Q31 are just a power amp mute.

So this circuit is there for power on muting then, correct? Right now even with the standard C64 1000uf cap in place there is a pretty good speaker thump at power on time.

Enzo

It is a mute, it shunts the INPUT signal to the power amp.   It can do absolutely nothing about the power amp's power up thump.

Roly, I THOUGHT it was you who mentioned something along those lines, but I see i was mistaken, I forget now which post was in my thinking.

txflood

Quote from: Enzo on March 26, 2013, 06:52:29 PM
It is a mute, it shunts the INPUT signal to the power amp.   It can do absolutely nothing about the power amp's power up thump.

Right, so the idea is that the transistor switch grounds the input (preamp output) going to the power amp until some time after the power amp is fully powered up. This would be governed by the RC time constants of the transistor circuit, correct? Just trying to learn.

Also I confirmed that the DSP board has 4 plastic standoffs. The only ground comes from pin 3 of the connector.

domohawk

#28
Sorry if it is bad form to revive a year old thread, but all the info is here and my findings directly build on the previous posts.

My GTX-65 has the same terrible "hum with dsp attached" problem as txflood and many others. I attempted perform the "switch J10's pin 1 to ground" as pondered in this thread, and while some high freq noise was cut (from dsp I assume) the main hum was still there.

I have since found that simply disconnecting J10's pin 4 (the power pin) removes the hum completely (as well as dsp air noise).

My current plan is to add a switch to break/join pin 4, however this creates a loud pop when disconnecting, which I suppose is not good. I do not mind only using this switch when the amp is off, but perhaps this gives some people in here some insight into the nature of the problem/a better fix?

Reference schematic: http://music-electronics-forum.com/attachments/6109d1251661835-195sch_2.pdf

Edit: This may also explain why over filtering at C64 also reduces the hum significantly, perhaps the dsp board is not well-behaved at its power input?

Roly

Very welcome by me.   :dbtu:


Ah HA!  Comparing this .pdf with GFX-120a circuit (<gx 212 gfx.pdf>, here already apparently), there is an interesting difference at pin 4 where the DSP power feed is.

In one circuit it is simply a 10 ohm series resistor, but in the other circuit they have lifted the resistor to 47 ohms and added an 0.1uF  bypass cap!  They have had a problem here.  I'd try adding this extra decoupling of the DSP power feed and see if it helps.
If you say theory and practice don't agree you haven't applied enough theory.