Topic: Rolls PB223 mod?

[Chuck]

OK, I just got one of these phantom power supplies and noticed that under load, it provides less than 30v DC with one MSH-1O microphone connected. So I got to thinking what it would take to beef it up a bit. I have attached the owners manual which includes the schematic for the unit. I know the schematic is for the PB224, but it matches the components in this unit too.

Anyway, I was thinking about changing some capacitors to larger values to start. I'm half thinking about putting a couple of transformers in it too, and removing the DC blocking caps to make it a pre-amp that provides phantom power also. There isn't too much room in there, but there may be enough room for some small transformers.

My initial though was to increase the values of C5, C11 and especially C6. My thought is that with larger capacitor values, it will have more reserve voltage and may be able to keep up with the voltage/ current needs of the microphones better.

If anyone else has any ideas on moding this, I'd like to hear them. I know there is another pre-amp project mod in this section. This may be an alternative to that, I think. Passive gain would be provided by the transformers and phantom power by the circuit in the PB223.

I will post photos as things go along.

Chuck


edit...
changed MSH-1A to MSH-1O... opps...

[poorlyconditioned]

OK, I just got one of these phantom power supplies and noticed that under load, it provides less than 30v DC with one MSH-1A microphone connected. So I got to thinking what it would take to beef it up a bit. I have attached the owners manual which includes the schematic for the unit. I know the schematic is for the PB224, but it matches the components in this unit too.

Anyway, I was thinking about changing some capacitors to larger values to start. I'm half thinking about putting a couple of transformers in it too, and removing the DC blocking caps to make it a pre-amp that provides phantom power also. There isn't too much room in there, but there may be enough room for some small transformers.

My initial though was to increase the values of C5, C11 and especially C6. My thought is that with larger capacitor values, it will have more reserve voltage and may be able to keep up with the voltage/ current needs of the microphones better.

If anyone else has any ideas on moding this, I'd like to hear them. I know there is another pre-amp project mod in this section. This may be an alternative to that, I think. Passive gain would be provided by the transformers and phantom power by the circuit in the PB223.

I will post photos as things go along.

Chuck

I'd like to see photos of the inside.

*But* before you go worrying about phantom/48V, check the PB223 with *no load*.  Remove the mic and measure the voltage, it should be 48V.  It will be less (like 30V) under load.  Why?  Well, there is a 6.8k resistor between the 48V supply and the mic connector.  The fact that it is 30V means that some current is flowing, and 18V is dropped on that 6.8k feed resistor.  Perfectly normal!

You could still put transformers in.  But be aware that you need a low impedance source to drive them.  The MSH mics are really a mid-impedance source, so there might not be that much gain.  An AT853Rx (AT853 with phantom adapters) would be perfect, though.  That, and most "professional grade" mics, will output low impedance suitable to drive a transformer.

  Richard

[Chuck]

Thanks for joining in Richard. I appreciate your input. I hope to have photos of the interior tomorrow. My digital camera works best with sunlight. 

This unit was busted when I got it. I think someone had reversed the voltage, because D1 1N4001 diode was burned. I replaced it and made a jumper to it, since the trace was also fried. It seems to work up to spec now.
The voltage is ~45v with no load. FWIW, the voltage on my PS-2 doesn't drop more than 3 volts with the same microphone.

One thing I don't understand in the circuit is the 7805. Why use a 5 volt regulator in a 12 volt circuit? It's weird, because with the switch set to 12v, it does supply 12 volts, even though the 7805 should only output 5 v DC.

As far as the impedence problem with the transformers, maybe I could make a small transistor buffer for the transformers?

[poorlyconditioned]

Thanks for joining in Richard. I appreciate your input. I hope to have photos of the interior tomorrow. My digital camera works best with sunlight. 

This unit was busted when I got it. I think someone had reversed the voltage, because D1 1N4001 diode was burned. I replaced it and made a jumper to it, since the trace was also fried. It seems to work up to spec now.
The voltage is ~45v with no load. FWIW, the voltage on my PS-2 doesn't drop more than 3 volts with the same microphone.

One thing I don't understand in the circuit is the 7805. Why use a 5 volt regulator in a 12 volt circuit? It's weird, because with the switch set to 12v, it does supply 12 volts, even though the 7805 should only output 5 v DC.

As far as the impedence problem with the transformers, maybe I could make a small transistor buffer for the transformers?

There may be other problems besides that diode you fixed, then.

As far as adding buffer transistors.  Yes, you can do that.  But at some point, it makes sense just to get a preamp.  Even a cheap preamp may do better than a buffer.  I'm pretty much done building preamps, since you can buy better quality, and there is less chance to screw up.  The problem is a lot of these circuits *sound* fine, but are still broken.  Look at the UA5 thread I started way back.  I noticed huge distortion (up to 1%) near 0dB input.  Yet everyone thinks it sounds great!  So, just because something sounds great, check it out with some test signals to make sure everything is fine.

  Richard

[Chuck]

I just got the PB223 yesterday, and with visual inspection determined that replacing D1 and the trace was prudent..I'll run a signal through it this weekend to see if it has more problems that need to be fixed.

I know what you mean about being able to buy good quality microphone pre-amps that have already been thought out. I figure I already have a good active pre-amp to use. I like to tinker, and a transformer based pre-amp appeals to me on several levels. My initial reason for getting the PB223 was to use it as the phantom power supply for an op-amp based pre-amp I made a while ago. Unfortunately that unit is built around a differential supply voltage based design and I decided I don't want to rework it for use with this phantom power supply after all.

The key thing for me would be to find some reasonably priced transformers (that will fit in the enclosure) to provide clean gain.

BTW, the newer MSH-1O's are more sensitive than the MSH-1A's and supposedly can handle higher SPL's. I have yet to test the 1O's for high SPL's, but I do like the sound of them and especially the price.

[Chuck]

PB223 photos:

[poorlyconditioned]

BTW, you know that Rolls has schematics on their website, right?

Every product that I've seen has a schematic included in the docs.

Not great products, but certainly solid, well made ... and well documented.

  Richard

[Chuck]

BTW, you know that Rolls has schematics on their website, right?

Every product that I've seen has a schematic included in the docs.

Not great products, but certainly solid, well made ... and well documented.

  Richard

Yes, I have the schematic for this unit as an attachment in my first post.
It's unusual for a company to publish the schematic like that.
I'm also thinking I want to build a high pass filter on this thing too.

[Chuck]

Modifications made so far:

Replaced non-functioning D1, 1N4001 with 1N4001
Added a jumper wire from D1 to ground as that trace was damaged
Replaced R4, R5, R10, R11 (5% - 6.8 k ohm) with 1 % - 6.81 k ohm metal film resistors (hand matched to .1%)
Replaced R7 (100 ohm) with a jumper wire
Replaced damaged U1, CD4069UBCN with CD4069UBE
Replaced damaged U2, 7805 with 7805
C1 (1 uf – 50v) added 100 uf 63v capacitor in parallel (old cap is under the circuit board)
C5 (1.5 uf – 50v) added 100 uf 63v capacitor in parallel (old cap is under the circuit board)
C6 (47 uf – 50v) added 68 uf 63v capacitor in parallel (old cap is under the circuit board)
C11 (1.5 uf – 50v) added 100 uf 63v capacitor in parallel (old cap is under the circuit board)
Removed SW28 and replaced with a jumper wire for to eliminate 12v DC option
Replaced buss wire grounding to SW28 with heavier wire and added a grounding screw on the chassis

With these modifications, I gained (3) v DC when using the MSH-1O's. Not a huge benefit. I already had everything but the CD4069UBE and the 6.81 k 1% resistors, so all it's cost me so far is $.39 or the IC and $.80 for 8 resistors, and some time.
I'm hoping to find some transformers that would fit in there to provide passive gain. I also want to make SW28 engage a high pass filter. Any ideas on a good choice for the high pass filter cut off and slope?

[rowjimmy]

Maybe I shoulda taken electronics instead of that Jazz history class...

I really wish I knew what all this meant.

[Todd R]

Excellent work, Chuck!

I'd make the hi-pass 75-80 Hz and -6db slope, a pretty good range for our applications.  If you're only planning on using this for the MSH omnis, you might want 100-120 and -6db.

[Chuck]

Excellent work, Chuck!

I'd make the hi-pass 75-80 Hz and -6db slope, a pretty good range for our applications.  If you're only planning on using this for the MSH omnis, you might want 100-120 and -6db.

Thanks, Todd.
I'll try both HPF suggestions and see what works best with the MSH-1O's.

Edit to add:

The 4069UB is a medium speed hex inverter, I wonder if substituting a different one, maybe a higher speed hex inverter would make a positive difference?

[Chuck]

Well, this weekend, I replaced U1 CD4069UBE (medium speed hex inverter) with CD74HCT04E (high speed hex inverter) and
replaced D8 (red LED) with (yellow LED). I gained another couple of volts DC under load on the phantom power using the high speed hex inverter. I replaced the red LED with yellow because everyone knows yellow is better. 

I also started to wire in some audio transformers. A photo of them is attached. I still don't know the manufacturer or specs, but my preliminary tests indicate that they are indeed impedance matching transformers and they end up supplying approx. 12 db of gain with the MSH-1O microphones.

I almost fried the phantom power circuit when I initially powered up the box after wiring in the transformers, because I connected the ground on both ends of the transformers to the chassis. Bad idea  So now, I'm looking for some isolated RCA panel mount jacks. I have some isolated 1/4" phono jacks, but no RCA's. It looks like I may have to put a little money into this thing now. So far it's cost me a little over one dollar in parts that I didn't have just lying around. If anyone has a line on some cheap good isolated RCA jacks, let me know.

[Chuck]

This is essential how I hooked up the transformer. Except I have tied pin 3 to pin 1 (Chassis ground) on the output, because I'm using RCA (unbalanced) outputs. The problem is this shuts down the phantom power circuit. I must be missing something that's very obvious. Of course, I'm at work, thinking about it, rather than having it in my hands to figure out. The inputs to the transformers (primary) are floating as in the diagram attached to Pins 2 and 3. As soon as I attach the chassis ground wire from the Phantom power circuit, it shuts that circuit down. That's what this diagram says, is to lift the chassis ground on the primary side. Also, If I connected Pin 3 to Pin 1 on the input side, that would produce a dead short for the phantom power supply. So that wouldn't work. If I leave the phantom power chassis ground disconnected, it all works, but with massive hum. Unfortunately, I won't have time to test anything tonight, since I'm working late.

[Chuck]

I've had some time to test and listen to this unit, now. The problem I was having with the transformers has been solved. There was a lead from one of the components touching the chassis. The transformers were wired correctly after all. After listening with the MSH-1O's attached, I noticed that the CD74HCT04E (high speed hex inverter) that I substituted was generating much more noise and grainyness than the CD4069UBE (medium speed hex inverter). So, I'm back to the original hex interter IC now. I think I've done about as much as I want to to the phantom power circuitry for now. I may end up putting in better bypass capacitors in the circuit, but for now I'm happy with the way it sounds. Next thing will be adding the high pass filter. I hope that adding more wiring and a switch doesn't increase the noise in the circuit. Essentially, the only real noise in this pre-amp is generated by the phantom power circuit, since the transformers (gain) are passive.

Comparing the stock PB223 to the Denecke PS-2 (which I also own) I have come to really respect the thought and especially the engineering that went into the PS-2. Especially, since they both use the exact same hex inverter to drive the cascading diode DC voltage gain circuit. The PB223 is a toy in comparisson. With both MSH-1O's connected to the PS-2, the voltage stays within 3 volts of 48 v DC. With two MSH-1O's connected to the modded PB223, I'm getting between 14v DC and 15v DC.

[Chuck]

So, here's where I'm at with this so far: (see attached .jpg drawing)

I plan to remove C3 and C4 as they are no longer needed. I don't like to see those electrolytic capacitors in my signal path either.
I'm going to build in a switchable high pass filter, using the existing existing 12v/ 48v DPDT switch.

I'm thinking that the logical place to add to single pole filter (-6 db per octave) is after the transformer. Though, I've seen it done on the microphone side too. If anyone has any thoughts on this, I'd really appreciate hearing them.

By replacing R9 with a smaller value resistor, I could make a 75 Hz to 125 Hz high pass filter using quality capacitors in the .1 - 2 uf range for C*. Since I don't know the impedance of the transformer output, I'll be guessing, with trial and error, on the exact values needed. I wonder how to acheive good quality control on capacitor values? A 10% fluctuation in value, could have a huge impact on the matching of the channels.

Again, I'd really appreciate any input, if anyone sees any problems or has suggestions on making this project better.

[Church-Audio]

So, here's where I'm at with this so far: (see attached .jpg drawing)

I plan to remove C3 and C4 as they are no longer needed. I don't like to see those electrolytic capacitors in my signal path either.
I'm going to build in a switchable high pass filter, using the existing existing 12v/ 48v DPDT switch.

I'm thinking that the logical place to add to single pole filter (-6 db per octave) is after the transformer. Though, I've seen it done on the microphone side too. If anyone has any thoughts on this, I'd really appreciate hearing them.

By replacing R9 with a smaller value resistor, I could make a 75 Hz to 125 Hz high pass filter using quality capacitors in the .1 - 2 uf range for C*. Since I don't know the impedance of the transformer output, I'll be guessing, with trial and error, on the exact values needed. I wonder how to acheive good quality control on capacitor values? A 10% fluctuation in value, could have a huge impact on the matching of the channels.

Again, I'd really appreciate any input, if anyone sees any problems or has suggestions on making this project better.

You need a capacitance meter to match them, you can also use software and a computer sound card and do a loop back into your computer to see the roll off of you hpf in action. Here is an example of one of my hpf designs in action.

This is the frequency response of my st-9100 preamp the red line is HPF on the blue line is with the HPF off with computer software you can see the roll off of your design ( this does not mean it will sound good your ears are the best judge of that. But it is a starting point, I can use this program to make sure left and right are exactly the same frequency response and if the caps are off value they will not be I hand measure all of my caps so that left and right are linear. A capacitance meter can be purchase for about $99 get a B&K they make great ones it really does not mater if its 100% accurate as long as the two caps are reading the same on your meter your golden.


Chris Church

[Chuck]

Unfortunately, I'm not inclined to buy a special meter just to measure the caps.
I must have half a dozen meters, but none measure capacitence.

[Church-Audio]

Unfortunately, I'm not inclined to buy a special meter just to measure the caps.
I must have half a dozen meters, but none measure capacitence.

I can measure them for you and send them back if you want. What valiues do you need? I might have them here.

[Chuck]

Unfortunately, I'm not inclined to buy a special meter just to measure the caps.
I must have half a dozen meters, but none measure capacitence.

I can measure them for you and send them back if you want. What valiues do you need? I might have them here.

Awesome offer, Chris... +T
Let me work some things out and I'll definately get back to you.

Do you think I'm on the right track here, adding the filter after the transformer rather than before it?
With a 10k resistor and a .15 uf cap, I get a corner of 106 Hz. Does that sound like a good place to strat? ToddR suggested between 70 Hz and 120 Hz with a single pole filter.

[Church-Audio]

Unfortunately, I'm not inclined to buy a special meter just to measure the caps.
I must have half a dozen meters, but none measure capacitence.

I can measure them for you and send them back if you want. What valiues do you need? I might have them here.

Awesome offer, Chris... +T
Let me work some things out and I'll definately get back to you.

Do you think I'm on the right track here, adding the filter after the transformer rather than before it?
With a 10k resistor and a .15 uf cap, I get a corner of 106 Hz. Does that sound like a good place to strat? ToddR suggested between 70 Hz and 120 Hz with a single pole filter.

I use a 330nf cap for a 10k input and that works well. I try and stay away from 100hz I find it eats to much of the warmth up. I like 80hz or even 70hz as a HPF when you get into 100hz your eating up the warm range of 150 to 250 as well these are critical areas. Sparkie is a good source for knowledge about cap values, I just change things until I like the sound of it because its about sound in the end, not just values.
It depends on a lot of things like the transformers your using and what the input impedance is. Simply strapping a 10k resistor to a cap although it does work might not sound great. You really need to just change the value of the coupling caps in the first stage of the preamp, I use a DPDT switch and "swap" the coupling caps out in the front end of my preamps. I find that is a better solution.

[Chuck]

So, you are suggesting not using the R/C circuit after the transformer, but rather just using the DPDT switch I have to engage a different value capacitor without the tie down resistor on the microphone side? Currently, the DC blocking caps are 47 uf electolytics. If I make (2) 330 nf capacitors switchable on Pins 2 ands 3, you think that would be the way to go? For some reason, I thought the tie down resistor was needed to correctly time the capacitors charging to the correct frequency. I'd need a quad throw switch to make that work, as that microphone side is balanced. I want to get rid of the 47 uf eletrolytics, as I think they are needed anymore, now that I have a transformer there to block the DC to the pre-amp side. .

[Chuck]

I finally, today, figured out what I'm going to do with this project 

I'm going to make it a three (maybe four) channel mixer!
I have always wanted a cheap/small/ easy way to mix several sources.

My new plan...

-Use the existing phantom power circuit to power a third (or fourth) microphone.
-Add in two line input jacks.
-Add a passive "blend" control to mix the extra microphones (brought up to line level with the transformers) with the line level source.

So, I could run my regular rig. Patch the line outputs on the PMD671 to the inputs on this new box. Then mix in a third microphone (typically an omni) for a three microphone mix. I'd then run this signal (line in) to a DAT.

This way, I'd have my usual two microphone recording and then a three (or possibly four microphone) mix as well.

The blend contol could be a simple 10k pot (tied to ground on one side) that feeds the microphone signal (at line level) equally to the line outputs.

To me, a device like this would actually be useful.

[Chuck]

Can someone tell me if using (4) 9.6 v NiMH rechargables is not as good as using (6) v SLAs in series to provide the cleanest phantom power? Is the chemical reaction that creates the DC cleaner with the SLA cerlls as opposed to the NiMH? I am really itching to get going on this project.

Also, is there any benefit in put large capacitance in parallel with the batteries, like used in automotive sound systems?

Thanks,

Chuck