Bogen MXM Tube preamp/mixer questions

[Potato Cakes]

Hello, everyone,

I am helping a friend with a Bogen MXM preamp/mixer that he is wanting to use for tracking. When he got it someone else had did some mods on it like replacing the filter switch with a pad and adding unbalanced direct outs on all of the channels which interrupts signal to the master. When tracking drums he said the inputs sounded overdriven which is fun on certain applications but not all. Since it's just going to be using as a mic preamp, I want reinstall the HPF and see about getting more headroom or padding the input to better accommodate recording drums. I've attached the schematic of the MXM-A which has a couple more features like the Hi-Z selectors, which this unit does not have. But the rest of it should be the same.

I was looking over the schematic and I noticed that the plate voltage to the 12AX7s on mic inputs 1-4 is 40V and I thought this was very low and that maybe this might be causing them to sound overloaded so easily if the plate is being starved of voltage. Would there a problem if I jumped R2, 8, 14, and 20 to get 185V for the plate like is shows for V3B?

I also noticed that there is no summing resistor network. Shouldn't there be one?

There is one problem that needs to be addressed before I get too far in the project. The master output has 4VDC between pin 3 and ground. There is also a very present 60 hum and the audio is severely high passed. The audio direct from each channel is correct, minus the headroom. The fun part is when I power the unit off it blasted my converter inputs with signal and I measure 53VDC and was dropping with the capacitors. I do not see how that is possible per the schematic. Granted I haven't begun to trace to see if the person whoever modded this unit missed something. It sounds almost like a bad transformer but I still don't know how DC is getting through. I may need to add an electrolytic inline to block DC but that won't fix the high pass issue. Any see something like this before?

Thanks!

Paul

 

[radardoug]

Thats a pretty Michael Mouse design, you either need to mod it significantly or not bother. The channels do mix by a passive method with the volume controls wired backwards.

 

[Potato Cakes]

I'm definitely not looking to overhaul this thing, just get the gain more usable to work with drums, which really would can be achieved using a pad at the inputs and call it good. It would be great to get the master section working but it's not critical.

The one thing I found is that the direct out mod picks off from the CW connection on the gain pot, making it unbalanced at the output. It seems to be fine as far as output is concerned but I don't know if it's the half of the 12AX7 is being loaded too much. From doing my own research it seem the dual tubes use one gain stage to feed the other for better headroom (I think) so just using half of a tube seems to be not as optimal. Again, I'm not rebuilding this guy. I'm just trying to make some fairly simple improvements.

I did think about the the volume pots creating the summing network but I didn't know if that would be a problem since when you change the volume.

Thanks!

Paul

 

[ruffrecords]

There is nothing you can really do to improve the output from a single 12AX7. its output impedance will be in excess of 50K no matter what you do and the direct out will be unbalanced because the whole of the mixer is internally unbalanced (which is true for just about every mixer. If you could wangle it a supply, about the best you could do would be to use an ac coupled TL072 to buffer each direct out and create a balanced output.

Do you have the mic input transformers fitted? Any pad you add should go before the transformer. You can buy 20dB pads built into an XLR which you just plug into the mixer input.

Cheers

ian

 

[Tubetec]

If you want to take outputs direct from the 12ax7 input stages they'll want to see hi-z (1megohm) active DI boxes might be your best bet . There seems to be a variety of octal transformers from Bogen available quite cheaply, might be worth having a few to try out . 100k ohm anode load resistors might work out better for relatively hot signals from drums close mic'd with dynamics . Id also go with resistor/cap bias
PSU is a bit skanky for a preamp too , halfwave ht rectification and series supplied heaters  :-\
No easy way to turn a sows ear into a silk purse unless the owner wants to throw money at it .

 

[Potato Cakes]

If you want to take outputs direct from the 12ax7 input stages they'll want to see hi-z (1megohm) active DI boxes might be your best bet . There seems to be a variety of octal transformers from Bogen available quite cheaply, might be worth having a few to try out . 100k ohm anode load resistors might work out better for relatively hot signals from drums close mic'd with dynamics . Id also go with resistor/cap bias
PSU is a bit skanky for a preamp too , halfwave ht rectification and series supplied heaters  :-\
No easy way to turn a sows ear into a silk purse unless the owner wants to throw money at it .

I'm going to add some of the A72DI PCBs that are offered in the White Market. Seem to have 1M impedance and there is a phantom power supply added to the unit which seems to be able to hand the draw of 5 direct boxes. Pretty inexpensive way to sort out the direct outputs.

I was looking at calculating the HPF capacitor value but I was not sure which resistor value to use. From looking at the schematic it seems that R41 (220k) is what is being used as the resistor to go with C4. But when I put those values in the corner frequency calculator it says that the HPF is 400Hz. Is this correct? It seems high for speech as it is labeled in the schematic.

I also noticed that the unit I working on is missing the capacitors before the tube sections for the mic preamps. Is there a good reason to omit them? I'm asking because I need to in a part order for a number of projects and I don't believe that I have the correct value to reinstall C1, 6, 11, and 16 if they are indeed needed.

Thanks!

Paul

 

[abbey road d enfer]

I was looking at calculating the HPF capacitor value but I was not sure which resistor value to use. From looking at the schematic it seems that R41 (220k) is what is being used as the resistor to go with C4.

The HPF is defined by C4 and the 500k pot. The actuak value depends on teh volume setting but stays pretty much stable, so teh actual corner frequency is about 170Hz.

I also noticed that the unit I working on is missing the capacitors before the tube sections for the mic preamps. Is there a good reason to omit them?

These caps are necessary only if DC is present on the source. Someone took them off thinking naively it would improve the LF response. You don't need them.
With the caps in place, the cut-off is at 4 Hz, FWIW.

 

[gridcurrent]

I also noticed that the unit I working on is missing the capacitors before the tube sections for the mic preamps. Is there a good reason to omit them? I'm asking because I need to in a part order for a number of projects and I don't believe that I have the correct value to reinstall C1, 6, 11, and 16 if they are indeed needed.

those capacitors are needed if you are retaining the grid leak bias of the input tubes.
if you want to omit them, revise the circuitry to cathode bias.

 

[abbey road d enfer]

those capacitors are needed if you are retaining the grid leak bias of the input tubes.
if you want to omit them, revise the circuitry to cathode bias.

OOps! I diddn't notice the lack of cathode resistor.
Thanks for reminding.
So, yes, these caps should be reinstalled. Actually, the plate voltage must be even less than 40V. Talk about starved...
No wonder it doesn't sound too good...

 

[Potato Cakes]

The HPF is defined by C4 and the 500k pot. The actuak value depends on teh volume setting but stays pretty much stable, so teh actual corner frequency is about 170Hz.
These caps are necessary only if DC is present on the source. Someone took them off thinking naively it would improve the LF response. You don't need them.
With the caps in place, the cut-off is at 4 Hz, FWIW.

I had it in my head that it had to be capacitor then resistor in the direction of the signal flow for HPF. I also did not mention that the direct out line as it is currently implemented is coming off of the wiper and not the CW connection of the 500k pot. In this instance, would the R41 (220k) now be used for HPF? I did run some tests and I am getting roll off about what I am calculating with 220k as the resistor value, but I didn't know if the variation of the potentiometer was also taken in a consideration. It may be negligible but I would at least like to know if I should be made aware.

I also want to emphasize that I do know that this circuit has a number of issues especially when trying to use the preamps directly, so it may seem like this endeavor is not the best use of time but ultimately it is knowledge that I am after so that I know what is happening when something like this comes up again.

Thanks!

Paul

 

[Potato Cakes]

OOps! I diddn't notice the lack of cathode resistor.
Thanks for reminding.
So, yes, these caps should be reinstalled. Actually, the plate voltage must be even less than 40V. Talk about starved...
No wonder it doesn't sound too good...

Would it be beneficial to remove R8 to get the voltage closer to 185V like it does for V3B? I had noted that the 40V seemed very low but didn't know if increasing the voltage would cause other problems.

Thanks!

Paul

 

[rackmonkey]

Back to your original question in the post, I didn't look closely at the input stage at first, but Gridcurrent noting that it's a grid leak bias setup triggered a light bulb. I was asked a while back to take an old Sony SRA-2 recording amplifier and make it into a more usable preamp. It had a grid leak front end as well, and had really low headroom (the main issue the owner wanted me to address). Converting it to cathode bias was a quick fix that made it into a nice little tube pre. Maybe try converting one channel and see how it goes.

Grid leak also has the liability of bias being less stable over time, as the tube ages.

 

[emrr]

Don't eliminate R2/etc, if anything lower the values of R4/R6/etc.  R2 down to 100K maybe.  Doubling voltage actually doesn't buy you that much more headroom in terms of absolute dB, 40 is a hair low but lots of the classics are in the 50-60V range. 

As said, pad needs to be a 20dB type in front of the transformer if needed.  Almost all old tube gear needs a 20dB pad most of the time with modern sources, which is problematic in that any times it reduces gain too much in a fixed gain amp situation like this. 

Definitely change it to cathode bypass and assess headroom then. Grid leak is usually pretty low headroom.  Consider leaving a cathode bypass cap out to lower gain some, definitely compare both ways.  Put cathode cap bypass switches for a +/-6dB option. 

The speech filter you mention is what they'd have done, it's what you need with close talked ribbons especially, and most other things through a PA back then. 

The DI approach is good.  I noted on a early build I did, had tried a 12AX7 into a 15K transformer and the gain loss from loading was something like 12dB.  It's not an output stage tube! 

I don't see anyone mentioning C3/R3/etc, that's a treble boost EQ that works by padding the lows, it's also there to give load isolation from the mixing pots.  If you're going to direct out only, dispense with that and pull signal from C2/etc.

 

[Potato Cakes]

As said, pad needs to be a 20dB type in front of the transformer if needed.  Almost all old tube gear needs a 20dB pad most of the time with modern sources, which is problematic in that any times it reduces gain too much in a fixed gain amp situation like this. 

Based off of what I was hearing and seeing when just running tone I wound up with a 30dB pad on the inputs. I tried 20dB first but I was getting too much harmonic distortion for the intended application.

[/quote]
Definitely change it to cathode bypass and assess headroom then. Grid leak is usually pretty low headroom.  Consider leaving a cathode bypass cap out to lower gain some, definitely compare both ways.  Put cathode cap bypass switches for a +/-6dB option.
[/quote]

For V1A/B and V2A/B the cathode goes directly to ground however on V3B (input 5 mic switch selected) there is a cathode bypass but it also has the 220k resistor omitted going to the plate making the plate voltage 185V, so my thought process was to copy and paste V3B to V1 and V2. If this not the case then I will request assistance on calculating the value of the cap needed and the resistor that would need to be added for a cathode bypass.

[/quote]
The DI approach is good.  I noted on a early build I did, had tried a 12AX7 into a 15K transformer and the gain loss from loading was something like 12dB.  It's not an output stage tube! 
[/quote]

Yes, indeed.

[/quote]
I don't see anyone mentioning C3/R3/etc, that's a treble boost EQ that works by padding the lows, it's also there to give load isolation from the mixing pots. If you're going to direct out only, dispense with that and pull signal from C2/etc.
[/quote]

I haven't figured out if the the mixer can be used yet as I haven't figured out how voltage is getting to pin 3 of of the output XLR which is a constant 4VDC when powered on and then spikes to over 60V when powered off and drains with the capacitors. For now I am going to leave on C3/R3 in case I do get the mixer section working properly as that was something that was discussed with my friend but we agreed if it can be salvaged then not to worry about it.

Thanks!

Paul

 

[Tubetec]

Socket X6 should have a transformer in it , seems like someone may have wired across the socket . If the transformer is gone you will need a high resistance path to ground to discharge any voltage that could appear , most likely though if you measure 4v dc with your meter at the output the cap C47 is leaky . Replace C47 with something modern and add a 10 meg resistor from output to ground , now you should no longer see DC at the output or get any loud thumps, the transformer will still be needed to feed lo-z balanced inputs via the xlr ,alternatively you could use another of your DI modules .

 

[Potato Cakes]

Socket X6 should have a transformer in it , seems like someone may have wired across the socket . If the transformer is gone you will need a high resistance path to ground to discharge any voltage that could appear , most likely though if you measure 4v dc with your meter at the output the cap C47 is leaky . Replace C47 with something modern and add a 10 meg resistor from output to ground , now you should no longer see DC at the output or get any loud thumps, the transformer will still be needed to feed lo-z balanced inputs via the xlr ,alternatively you could use another of your DI modules .

There is a transformer at X6 and the the schematic has a film capacitor inline from the last tube stage to the output transformer, and I thought film caps wouldn't have leaking issues.

I'm going get back to working on this later tonight. I may discover more alterations to the circuit than I previously found.

Thanks!

Paul

 

[Tubetec]

A transformer should block dc in any case , maybe its an autoformer/tapped choke .
Even film caps can become leaky .

 

[Potato Cakes]

A transformer should block dc in any case , maybe its an autoformer/tapped choke .
Even film caps can become leaky .

It's labeled as a line matching transformer on the chassis and is the last thing on the way out to the XLR, but what it actually is or how it's wound is a mystery as that info is not printed on the transformer nor have I found any information on it yet.

Thanks!

Paul

 

[gridcurrent]

For V1A/B and V2A/B the cathode goes directly to ground however on V3B (input 5 mic switch selected) there is a cathode bypass but it also has the 220k resistor omitted going to the plate making the plate voltage 185V, so my thought process was to copy and paste V3B to V1 and V2. If this not the case then I will request assistance on calculating the value of the cap needed and the resistor that would need to be added for a cathode bypass.

the plate, cathode and cathode bypass values associated withe V4a are ideal for the input stages,
and close to those suggested in RCA Receiving Tube manuals.

 

[Potato Cakes]

the plate, cathode and cathode bypass values associated withe V4a are ideal for the input stages,
and close to those suggested in RCA Receiving Tube manuals.

Thanks for the transformer info! I must be bad at the internets.

I have some parts to order then I will report back with my successes or failures

Thanks!

Paul