DMP3 - Schematics and mods....

evm1024

evm1024

New member
Well this is the first installment in the DMP3 mods project. As you may recall Antichef gave away his DMP3 and I was the one to win it (Super thanks again!). I wanted one to take apart and see what could be done to make a great pre better.

The first step has been to get a schematic. Of course none are available to be found on the net (at least for me) so as expected I traced out the circuit visually and hand drew it.

Here is a link documenting the effort.

http://arafel.org/audio/diy/DMP3/

So far we have a parts list and the schematics for the INA163 input amp, the gain selector / 75 Hz hi pass filter and the output stage.

Later I'll add schematics for the meter section and the power supply. In addition a little commentary on the circuits.

Lots of thought for mods to explore. Changing the PS is "easy" if you don't mind changing the case form factor. Other options are to change the 75 Hz (72 Hz actually) hi pass filter to be a complex filter to get a more colored sound. Then you could just change out some parts and drop some to try to get the most gain-on-a-wire. lastly changing the meter ro give more VU like results might be useful.

Regards, Ethan

PS if anyone want to to put this into a schematic capture program please feel free (and copy me)

Here is the input stage for a sample (make that a TRS rather than a TS jack)
 

Attachments

  • dmp3-input.jpg
    dmp3-input.jpg
    45.2 KB · Views: 1,053
Nice!!! I plan to do something to my DMP2s, and I reckon a lot of your work will be portable, since the circuits are similar.

Keep this up, and I'll be feeling like I should have paid you to take it :D
 
Questions

OK Here are a few thoughts on mods that I would like to hear your input on.

1) does anybody care if the 75 Hz filter goes away?

2) How about losing the Direct In?

3) would you like it if the gain range went up to 70 or 75 dB (for ribbons) This would also cause the lowest gain settings to go up 5 or 10 dB also.

4) what do you think about putting it into a new box (or how about 2 of them for 4 channels) with an outboard power supply (real 48 volts and great regulators)?

Just some food for thought. What do you think?

Regards, Ethan
 
Regarding #1, what does the 75Hz filter do??

#3...if this is going to be a very nice sounding pre, which I gather it will, it would seem to be a good partner for ribbon mics, so I think this seems like a good idea regardless of the equivalent boost at the lowest gain setting. Can't see why that would hurt anything and it would be great for it to be "ribbon ready".

Not that I ought to comment on any of this since I don't own one and don't know if I ever will, but I like drop in mods, so, related to #4, if I had a vested interest I'd be rooting for keeping it in the stock housing. Open it up, change components and whatever else, button it back up and listen to the sweet, sweet sound.
 
I continue to not get this project at all. It just seems like modding for the sake of modding. If what you want is a 4 channel instrumentation amp in a 1u box, with top quality parts, +70dB of gain, and a proper power supply, go here:

http://fivefishstudios.com/index.php?option=com_frontpage&Itemid=1

And buy his kit. By the time you are done with desoldering and metal fab, and etching the PSU PCB, and sourcing all the parts, you'll wish you had in the first place.

And you can even get 2 channels of his new trafo input pre for variety.


Anyway, my feeling on ribbons and gain are well known, I think. This is a quiet pre, but it still only needs 40dB or so to get above the input noise of a decent converter. 60dB is way more than enough. Tape decks, I dunno, maybe in that case. But I sell a LOT of inline pads, which tells me that minimum gain is a much bigger problem than maximum gain. Why that is I don't know, I never have a problem with either one. My guess is there are a lot of pres out there with significant minimum gain (and also a lot of people with APIs whose main goal in life is to crush the output trafo).
 
mshilarious said:
I continue to not get this project at all. It just seems like modding for the sake of modding. If what you want is a 4 channel instrumentation amp in a 1u box, with top quality parts, +70dB of gain, and a proper power supply, go here:

http://fivefishstudios.com/index.php?option=com_frontpage&Itemid=1

And buy his kit. By the time you are done with desoldering and metal fab, and etching the PSU PCB, and sourcing all the parts, you'll wish you had in the first place.

And you can even get 2 channels of his new trafo input pre for variety.


Anyway, my feeling on ribbons and gain are well known, I think. This is a quiet pre, but it still only needs 40dB or so to get above the input noise of a decent converter. 60dB is way more than enough. Tape decks, I dunno, maybe in that case. But I sell a LOT of inline pads, which tells me that minimum gain is a much bigger problem than maximum gain. Why that is I don't know, I never have a problem with either one. My guess is there are a lot of pres out there with significant minimum gain (and also a lot of people with APIs whose main goal in life is to crush the output trafo).
Click to expand...


Thanks for your input. Yes, this is modding for the sake of modding. Many understand that and are OK with it. I hope you will let us have our little folly. :D

In doing some of the research (hey wait that might be a reason too do this as well!) I did look at his kits.

Let's see, take a DMP3 that you've out grown or buy for less than $125, Add $25 (oh heck let's splurge add in $50) in parts and have a (marginally?) better pre than stock with the glow of Modding.

And the 5 fish runs $489.00 for 2 channel kits + PSU + Power Transformer + 1u Rack Case


Barring the Cinemag I would expect all ina163 like preamps to sound very much like each other. Why spend $500 when you can spend $200?

Why would you fab up a power supply when you can buy one form 5fish or JLM?


Regards, Ethan
 
evm1024 said:
Barring the Cinemag I would expect all ina163 like preamps to sound very much like each other. Why spend $500 when you can spend $200?

Why would you fab up a power supply when you can buy one form 5fish or JLM?


Regards, Ethan
Click to expand...

The 5F rack case is for four units, so it's not quite apples to oranges there, and you've left out the major cost of the rack case (with fab) from the DMP3 side. And I don't value my time at zero. Other than the incredibly hard-to-look-at red color he's chosen for his latest PCB batch, I could pop together 4 of his channels in probably an hour. If he used a pot rather than the rotary switch, probably a half hour . . . most wouldn't be that fast, but I do a lot of soldering. How long would it take me to fab a chassis, yank off the bits from the DMP3 x2 and stuff in new bits, mount in the case, etc.? That depends on how good the instructions are, I guess . . . ;)

Would I buy any of those kits or mod a DMP3 myself? No, I wouldn't. I can design my own circuits and get the PCBs very cheap, because I order PCBs all the time so I can bundle them together and keep costs low.

I mean, an instrumentation amp is a datasheet pre with a handful of components. My cost for a kit similar to 5F (minus the rotary switch, the biggest item by far, and the DC servo as caps don't give me heartburn) would be about $10 plus case and connectors. Say $100 for four channels, $120 for six, $140 for eight. I would use a chassis mount pot, so I wouldn't need fancy or accurate fab, just a few holes here and there. I wouldn't use XLRs, so my step bits wouldn't get too hot--it has to get wired into the patchbay anyway. I would do a +/-15V supply and use the +15V rail for phantom without a multiplier, because I don't use any +48V mics. I don't need meters because the ones that count are on my converters.

That's just me, my life is simple . . .
 
Hell yeah -- modding for the sake of modding!!! :D:D
evm1024 said:
1) does anybody care if the 75 Hz filter goes away?
Click to expand...
not me
evm1024 said:
2) How about losing the Direct In?
Click to expand...
That's OK too, I suppose
evm1024 said:
3) would you like it if the gain range went up to 70 or 75 dB (for ribbons) This would also cause the lowest gain settings to go up 5 or 10 dB also.
Click to expand...
I'm lost on this issue -- I have some cheap ribbons, and the DMP3 (that one, in fact :) ) had plenty enough gain for me -- I'm not "analog only", though. Anyway, I don't need any more gain, personally :D
evm1024 said:
4) what do you think about putting it into a new box (or how about 2 of them for 4 channels) with an outboard power supply (real 48 volts and great regulators)?
Click to expand...
That's what I want to do with my 2 DMP2s - I've got the box (a rack mount "power distribution unit"), and I'll have to use an outboard power supply, because there's no room for it in the box. Real 48v would be awesome!

For me, too, it'll actually be more than just about modding -- 1st, I'm *learning*, which is something I could do with a kit, I guess, but this sort of activity is working, and 2nd, I would really use the 4 channel pre, and I've already got the two DMP2s ($90 for both of them) and the rack box ("free" with my Digital MPA), so my total cost will be much less than a kit, and my incremental cost will be even less.
 
Can you modify the circuit, put a Jensen in it, and make it sound like a Hardy? :D

Oh, and for minimal cost...

I would like that a lot.:)

I don't need more gain.

I don't need the DI.

I don't use the meters.

I wouldn't want a new chassis.

Just better sound.
 
Only Hardy souund Like Hardy

Actually I have a twin servo that I'm building on the bench. That is where my Jensen JT-16 will go. I might hook them up to the DMP3 just to see what they sound like.:D

Building a MPC-1 from scratch is not too hard. John Hardy's site has lots of information. However, John puts more than parts into his preamps. He puts many many years of experience with these circuits and hand tunes them to near perfection. (I would like very much to have a few JH preamps!)

---------

This mod on the DMP3 is really about finding the compromises in the circuit and see what happens when they are taken out. The 75 Hz rumble filter for example puts a voltage divider in the audio path even when turned off. The DI puts 2 electrolytics in the audio path, Yikes.

I'll have a block of time between now and the new year. I'll be making the input section more closely match the data sheet (drop the DI and extra cap plus up the value of the caps) tune the input bias resistors for lowest noise.

Take the low pass filter out and drop the voltage divider. Change the opamps out to something else, I'll try a few different ones. Take out one more electro in the signal path and try a few DC offset compensation schemes (One from Hardy preamps and the other a servo).

Change out a few resistors for metal film and some caps for polypro film.

Then do a few recordings of an unmodded DMP3 next to the modded one for all to hear.

Thats what I'm thinking anyway.

Regards, Ethan

PS if were lucky we might get some better sound for a few bucks!

PPS I do think about making a 990 based pre with Cinemag iron. Might be a killer for $200 a channel.
 
Well, that is the wrong way to do a DI. The input impedance is still too low. They should have split off the DI to a separate opamp and have that feed INA163.

I don't think that will change the noise performance much though. And I wouldn't go too low on R4-R5 as it appears it is using the phantom resistors for setting mic input impedance, although you didn't trace that far up. In fact I would keep it high enough so that you can lose both electros and exchange for a film. Certainly 5uF is overkill at 100K.

I would just lose the low cut entirely and save modding the larger part of the circuit. Maybe put in a switchable cap on the input for a first-order filter, which always works fine for me.
 
mshilarious said:
...
I mean, an instrumentation amp is a datasheet pre with a handful of components. My cost for a kit similar to 5F (minus the rotary switch, the biggest item by far, and the DC servo as caps don't give me heartburn) would be about $10 plus case and connectors. Say $100 for four channels, $120 for six, $140 for eight. I would use a chassis mount pot, so I wouldn't need fancy or accurate fab, just a few holes here and there. I wouldn't use XLRs, so my step bits wouldn't get too hot--it has to get wired into the patchbay anyway. I would do a +/-15V supply and use the +15V rail for phantom without a multiplier, because I don't use any +48V mics. I don't need meters because the ones that count are on my converters.

That's just me, my life is simple . . .
Click to expand...

Very well put. I'm doing exactly this with INA217s and a FiveFish power supply (except that I'm putting XLR inputs on it, only because I may need to use it out of my studio and without a patchbay :) ).

I think I should be able to get 8 channels of very clean pres with unbalanced outputs for maybe $160 all-inclusive. Maybe a little more.

Brandon
 
Run on batteries

I would like to use it sometimes on my Zoom H4n for some very quiet outdoors ambients.

The specs show a consumption of 9 Watts, wich would give about 3 hours with 7 AA accus, but I suppose I can save a bit with replacing the bulbs of the Vu-meters with leds.

Did anyone ever try to run the DMP3 on batteries?
 
Last edited:
paindur said:
I would like to use it sometimes on my Zoom H4n for some very quiet outdoors ambients.

The specs show a consumption of 9 Watts, wich would give about 3 hours with 7 AA accus, but I suppose I can save a bit with replacing the bulbs of the Vu-meters with leds.

Did anyone ever try to run the DMP3 on batteries?
Click to expand...

I dunno, 9W is an insane amount of power for a portable device. Consider that INA163 only wants 10mA, times two that's 20mA. Where is the other 980mA going? Well, let's see actually we have to step up the 9V to 30V, so really we'll say 40mA, plus about 100mA for phantom . . . still a lot of missing power . . . but the unit has to be rated for maximum load at maximum level, that could chew another 400mA I guess . . .

But the bigger problem you will have is the DMP3 wants 9VAC, not 9VDC. This is because it probably has internal transformers to step up from 9VAC to the +/-15V and 48V it needs. To do that with batteries you'd need an inverter, or you'd need to replace the internal supply with DC converters.

Here is something to think about for portable use: whatever headroom you don't need costs you power. I'm going to guess the H4N is a -10dBV device. That means it would be happy with maybe +8dBV maximum input (probably less, but let's leave some headroom). That's about +/-4V supply. If you run +/-15V instead, that means your quiescent power is 0.3W per amp instead of 0.08W. That's also the difference between running direct from a 9V battery vs. requiring a DC converter, which will have probably 20% loss of its own.
 
Thank you very much for your answer, mshilarious.

I will still order it, but what a disappointment !

My skills are very basic, so I'm not sure I understood your answer.

About replacing the internal supply with DC converters, do you think I have a chance to achieve that without adding extra noise ?

Maybe I could try with the smallest UPS available, it would be the easiest, although not the most convenient.

I'm very interested in the last part of your answer, but I think I don't understand it :o

Does it mean I will have a bad result ? Can I fix it with an attenuator ?

I chosed that preamp because people say it's neutral, and specially because it's noise floor is low. I used to have a Nagra, and when I compare my old recordings in silent ambients with the ones of the H4n, I'm a bit frustrated.
 
Hey, I'm still watching for the rest of that schematic!

You missed a couple of resistors, too. There's a pair of I think 2k resistors on the XLR side of the jack switch, thus providing 2k ohms input impedance through the XLR and 100k ohms into the 1/4" jack when the switches open.

Check me on that. I was wrong once.

A
 
Ethan, did you ever finish your DMP3 meter schematic? My DMP3 meter lamps aren't getting voltage and I'm wondering what the culprit might be.

Thanks!

Otto
 
evm1024 said:
Well this is the first installment in the DMP3 mods project. As you may recall Antichef gave away his DMP3 and I was the one to win it (Super thanks again!). I wanted one to take apart and see what could be done to make a great pre better.

The first step has been to get a schematic. Of course none are available to be found on the net (at least for me) so as expected I traced out the circuit visually and hand drew it.

Here is a link documenting the effort.

404 Not Found

So far we have a parts list and the schematics for the INA163 input amp, the gain selector / 75 Hz hi pass filter and the output stage.

Later I'll add schematics for the meter section and the power supply. In addition a little commentary on the circuits.

Lots of thought for mods to explore. Changing the PS is "easy" if you don't mind changing the case form factor. Other options are to change the 75 Hz (72 Hz actually) hi pass filter to be a complex filter to get a more colored sound. Then you could just change out some parts and drop some to try to get the most gain-on-a-wire. lastly changing the meter ro give more VU like results might be useful.

Regards, Ethan

PS if anyone want to to put this into a schematic capture program please feel free (and copy me)

Here is the input stage for a sample (make that a TRS rather than a TS jack)
Click to expand...

Do you by chance still have a link to these mods? I have some lying around and wouldn't mind rehousing and modding them
 
You must log in or register to reply here.
Back
Top