Repair Carvin DI design defect

[Axtman]

I am repairing a Carvin Red Line R1000 amp that blows the DI op amp when phantom voltage is turned on.

Attached are snippets of the DI output sections of the schematics from Revision H (my amp) and Revision I (newer version). In doing my research it seems these amps went through many configurations (at least 14 since they stopped at Revision O).

Before I start cutting traces and bodging on components please review the schematics and let me know your comments. Thanks.

[g1]

That is often what happens when applying phantom to a DI. It's not really a design defect anymore than a mixer that doesn't allow turning off phantom independently for each channel is a design flaw. They are both design shortcomings I'd agree.
I'd consider it more of an update.
Yes, just a resistor and cap to add in series with each of pin 2 and 3 circuits.

[nickb]

If you've got phantom power on it then you must be plugging it into a the mic input. This DI should be plugged into the line input. You may need an XLR to TRS jack lead as many mixers use a jack socket for line in.

If you are determined to use the mic in then try adding one diode to each of the signal lines at the junction of the 150 ohm/ 22uf cap i.e two diodes. Anode to signal and cathode to +15 rail.

[Axtman]

I picked up this amp really cheap because of the DI problem. The seller informed me of the issue and even had the XLR output plugged with warning message.

Many years ago I had a similar issue with a similar Carvin amp where plugging the DI from the amp into the board would screw up the volume of the amp.

If I have to go direct into the mixer I use a separate passive DI box. I have found through my years of playing that DIs from amps tend to be noisy (and problematic).

[Axtman]

Do I need to use 50 volt capacitors? Since I have minimal room, I have to bodge them in and want to keep components as small as possible?

Has anyone had luck in cutting a trace, drilling a small hole in the traces, scraping back the solder masking and soldering a components in? I have repaired broken traces but never soldered a component in the middle of a trace. Any tips would be appreciated.

[Leo_Gnardo]

Do I need to use 50 volt capacitors? Since I have minimal room, I have to bodge them in and want to keep components as small as possible?

Has anyone had luck in cutting a trace, drilling a small hole in the traces, scraping back the solder masking and soldering a components in? I have repaired broken traces but never soldered a component in the middle of a trace. Any tips would be appreciated.

You have to block 48V, so a 50V rating on the caps is just barely adequate. Anything less is courting disaster. You've had enough of that, right? 63V rating if you can manage it. One option - use smaller value caps, at the expense of low bass response.

And yes I've done the cut & insert operation on pc boards. You might be able to skip the hole drilling if you can mount the caps on the trace side of the board. Imagine you're a brain surgeon, or at least a dentist, for a while... and bodge on!

[nevetslab]

I'd use 63V rating on the 22uF cap. Size difference between the 50V and 63V may not be different at all.

On adding holes to a PCB for installing this sort of mod, sure. I'd drill #58 (0.042" dia with a Dremel or Foredom hand-piece) adjacent to the circuit trace, cutting the trace between source (op amp outputs) and destination (XLR connections). Scrape away the solder mask, install the resistor lead, part standing up Asian-style, fold the lead over and tack-solder to the prepared trace. Do the same with the cap (axial lead cap would do well here). Put the Plus side facing the XLR pins. Fold the lead over & tack-solder to the prepared trace. Then, join the two leads in-air...resistor to the cap. If it needs it, apply a little RTV to hold the parts to the board from bouncing around in transit.

[J M Fahey]

Just a week ago we were discussing this exact same problem, in that case on an SWR amplifier, so please search and read it.
I copypasted the full schematic including diodes and all.

FWIW the resistor should be between the diodes and the XLR pins , so it can limit voltage peaks coming from Ļ*outside*.

And yes, you need 50V caps *at least* , next higher being much safer, with positive pointing towards XLR pins.
But a 1uF x 50V electrolytic is quite small and fits anywhere.
Make it 63V and weīre talking.

[Axtman]

J M Fahey,

Would you mind copying the link to your post? This forum will not let me search 3-letter words (SWR) and I can't seem to find the post.

Thanks!

[eschertron]

https://music-electronics-forum.com/...ad.php?t=47478

Here is the thread. Good read, BTW!

[Axtman]

Thanks eschertron and J M Fahey!

[nickb]

You have to block 48V, so a 50V rating on the caps is just barely adequate. Anything less is courting disaster. You've had enough of that, right? 63V rating if you can manage it. One option - use smaller value caps, at the expense of low bass response.

And yes I've done the cut & insert operation on pc boards. You might be able to skip the hole drilling if you can mount the caps on the trace side of the board. Imagine you're a brain surgeon, or at least a dentist, for a while... and bodge on!

This is a valid point and begs another question. In he schematic shown the caps as non-polarized but I wonder if electrolytics are fitted. If they are and negative goes to the XLR then they will fail when phantom is applied.

[J M Fahey]

They are definitely electrolytics, schematics show polarity, and in any case I already wrote:

with positive pointing towards XLR pins.

SWR already has caps in series with XLR outputs to avoid DC seeping in, and they point the + terminal "outside" because they expect a positive one, typically 48V (so caps should at lest stand 50V and even better 63V)

Schematic:


check C164/165

What you do not need is R21/22 , which they added so they can mute XLR outs, but

only need 2 cheap diodes per pin (#2 and #3 in an XLR) tack soldered to pads between Op Amp inputs or outputs and nearby +/-15V rails feeding Op Amps.

If you donīt currently have series resistors or capacitors, drill through tracks (if space allows) , cut them wherever needed, and I would add a drop of glue to hold components steady, a drilled track is quite weaker than a proper pad so we donīt want component weight flexing and tearing them when amp travels in the back of a pickup truck on a bumpy road ... serious.

Or even the typical Bass cabinet vibration.

I am considering ordering yet another punching/pressing die set to make "Ampeg style" metal cabinet "cups" because my heads are "walking" over cabinet tops and you have to catch them before falling, thatīs how much cabinets vibrate.

[g1]

This is a valid point and begs another question. In he schematic shown the caps as non-polarized but I wonder if electrolytics are fitted. If they are and negative goes to the XLR then they will fail when phantom is applied.

They are definitely electrolytics, schematics show polarity, and in any case I already wrote:

I think nickb was asking about the updated Carvin circuit:

[J M Fahey]

Ok.
Schematic shows no "+" sign so *maybe* they are Bipolar electrolytics, some designers use them a lot, such as old VOX or very early Acoustic, most do not, just use standard ones and rely on very little voltage dropping across them.

Now in an XLR out only DC polarity which can often come from outside is positive.