Bruel & Kjaer 2609 Measuring Amplifier

[SyncTronX]

Hi Gentlemen,

Perhaps you can assist with a recommendation for changing the Negative
output of the LM723CH in power supply circuit for a Bruel & Kjaer 2609
Measurement Amplifier.. Schematic attached.

I started the calibration and adjustment procedure

The negative rail is at -17.126 and it needs to be boosted to
the -18V. The exact specification of V20 Pin 5 is -17.7 to18.5 Volts.

There is barely enough room to change a resistor. It's a very tight
fit. Otherwise I would place a pot in it in the voltage divider with
the feed to the Inverting input. I could calculate it a resistor value but
I don't how how the LM 723CH responds to changes in Inverting voltage
changes if at all. I'd rather not have to pull and replace parts more than
necessary it if VERY tight right there.

I thought about doing it on the trace side of the board but there is no clearance
when the bottom is slid back on.

I've looked at the data sheet from TI and it doesn't help as their circuit
designs are different. Also puzzling on the schematic (Fig 18, Page 9) shows one extra pin for the 10 pin device 11 pins?

Here's the TI Link. LM723CH (rev C)

Any Ideas, suggestions?

Cheers,

Sync

 

[1audio]

I would disconnect the load first to make sure the regulator is operating correctly. It's unusual for a problem like this to surface from the regulator alone. Also check the resistor values in the circuit. They may well have drifted over the years. The only critical voltage is the bias for the Microphone. Everything else has feedback stabilization.
It is old so parts may have shifted. Does it operate correctly?

Sent from my SGH-M919 using Tapatalk

 

[Audio1Man]

I always check the INPUT POWER and shoot the parts with IR gun first. R64 & or R68 have changed in value. Many designers may overlook the inrush power on SMD. They don't have much thermal mass and will change value. I have had some analog circuits that the THD% was good @ 1kHz and 20dB worse @ 20Hz even with 25ppm parts and they were running @ 50% power.

If you need to adjust the voltage you may try to add / stack another SMD on top of the resistor to bring up the voltage.

As Demian said is it working correctly?

Good hunting
Duke

 

[SyncTronX]

Hi Demian,

I didn't have a load connected and I'm not sure which component would
be the load? Oh I got you. I think there may be an -18V bus, I'll have to
find it and test again without the load. I was thinking it should compensate
itself for the load.

It appears to work but there are some questions that I have regarding it.
The lights blew out on it over voltage?

The resistors in the immediate circuit appear to have drifted low?

Looking at the schematic I thought (C7) might be leaking and would
offset the negative voltage rail. C7 originally measured 29VDC with .363VAC
leakage across it. So lets replace and see the affect.

C7 was replaced with several different units, shown as:
PB, bipolar Nichicon - worst leakage in VAC .459
PET Nichicon - worse leakage VAC .394
a Chemicom - best leakage VAC .257, brought voltage to 29.7VDC.

Thinking the leakage would offset the negative voltage was false.
Interestingly it had no effect on -18 VDC voltage, but did raise that
b+ range to 29.7 VDC for VC and V+.

My worry is if I start disconnecting other resistors, I won't be able to get them back in
again without disassembling the whole measuring amplifier...and on
and on it would go. I don't have a reference that I know of that discusses
the operation on the LM723ch. Perhaps there is a design document somewhere.

I'll try calculating out the measured voltages given and compare against the
the formula on the TI data sheet, data sheet page 8.

Cheers,

Sync

 

[SyncTronX]

@Audio1Man,

The unit is an older through hole discreet component built amplifier.
I checked the power supply. The voltage in the center transformer
is w/in spec as shown on my schematic attachment.
23.4 VAC measured vs. 21V schematic.

The Vc, V+ rose to a measured 29.7 VDC input into the LM723CH.

Same with the Top transformer which feeds the 7818 regulator, its
output is right at 18 Volts. So that part is good.

The Pol. Voltage can be switched and adjusted exactly with it's pot
from 0 V, 28V, and 200 Volts. So from that perspective that works.

The hi power transformer winding is 219 VAC and at the bridge
the voltage is 293V VDC.

It would have been nice if there were other voltages in the schematic or
the manual but there aren't a lot.

But you did give me an idea, I can try to hack on some higher value
resistors across the existing ones on the board and see what that does
to the -18V line. Thinking just try to lower the resistance a little bit
and see what that does. I cannot raise the resistance by adding a parallel
resistor and I don't have room for a series resistor or adding a pot.

I guess I can start cutting some traces and adding series resistance that
way. First I'll see what lowering the resistances will do. It should make
the LM723CH regulator less negative.

Thank's for helping me think through this.

Sync

 

[Audio1Man]

Sync
Try adding about 70k across the 3k57 this should increase the -17v closer to -18v. The thing that you are calling leakage is ripple voltage. The Electrolytic caps can be +20 to -50% in value, so the slight change in ac ripple voltage with your different caps is normal.

I hope this helps resolve your problems.
Duke

 

[SyncTronX]

@Analog1man.

Thanks,
I'll give that a shot. Shouldn't take long to tack that across.
I also just finished up the TI calculations which didn't work
for this configuration. As measured values input into formula
gave me .947V out.

Vout = ([Vref/2] * [R1 + R2/R1])
.947 = [.8V/2] * [3.14k + 4.3k/3.14k]

Cheers,

Sync

 

[1audio]

I have 2 of those. They are really useful, especially if you have a B&k Mike preamp. I'll look inside of mine later. Is it working otherwise? The box opens like a Chinese puzzle box.

Sent from my SGH-M919 using Tapatalk

 

[1audio]

The LM723 and it's variants have been used in power forever. Look up ua723 for the old Fairchild version. Lambda used them liberally.

Sent from my SGH-M919 using Tapatalk

 

[SyncTronX]

Update:

@Duke,
Well that was quick. It ended up using a 39.2k ohm resistor in parallel
with the 3k57 resistor. That gave me -18.06VDC So that is a good thing.

@Demian,

So, does it work?

Parts of it work. Other parts do not. I am not sure what I am doing yet.

For example, after adjusting the 50mV internal reference to the meter
reference mark it is still right on the mark.

The DC output, well that is variable.
The AC output, that is variable also with the AC input.

But they don't track properly.

The DC output is supposed to end up at 3.16V, when the front
panel meter indicates 10V. That is the full scale indication. However,
the DC output that corresponds to that is about 4.6VDC

Maybe that can be adjusted. I'll keep working on it and see what I come
up with.

I don't have any mics that I can use with it and I have very large impedance
mis-match with the other instruments that I have. 2M imput. The generators
I have impedance of 600 ohm, 130 ohm, 75 ohm, 50 ohm etc.

Thinking it will be a challenge to find the 7V fuse looking lamps but
maybe not.

I thought that the 723 might be common, but now I know where to look
for it. Thanks.

I will get there.

Cheers,

Sync

 

[SyncTronX]

Looked for the old data sheet version of the 723 and can't find one that doesn't have
the 14 pin layout with the extra Vz that is used for negative voltage.

 

[1audio]

I think this will help http://www.repeater-builder.com/astron/pdf/understanding-and-using-the-723-vr-hr-89-march.pdf along with http://www.ti.com/lit/ds/symlink/lm723.pdf

If you have a scan of the manual I would appreciate a copy.

The levels in the measuring amps are all "relative". Even the meter scales swap. I have something like 8 different ones. The idea is to look up the sensitivity using the 50 mV or use a calibrator and turn the screw until the correct meter reading is found. The voltage out is a similar relative reading. The 2709 does not have the log converter that the other measuring amps have or some of the other confusers that make them hard to use. Its very different from our concept of a normal meter.

I fired up one of mine. Short form info:
1) response rolls off about 3 dB on meter and AC out at 100 KHz
2) Internal noise limits the distortion, ranging from .03% to .5% depending on range and level. The actual distortion is lower, around -80dB for the harmonics.
3) If you set the meter for FS at 100 mV you get 3.16V RMS out.

Its not particularly good SINAD or THD+N, limiting some measurements. I gave up on the B&K stuff for distortion measurements of transducers since the electronics limited the measurements. I use some very vintage HP stuff where the mike preamp to output THD is .005%. Partly why I get much lower distortion on things like headphones than the vendors I work with.