开云体育

Hey y'all! I am reaching out regarding an SG503 leveled sine wave generator I am trying to repair. It's mostly in good shape, however, the 7-segment display does not work - it just displays all zeros regardless of what the frequency setting is. I started by looking at the -22V and 5.2V rails. Ripple is very low for both, however, they are both low (-20V and 4.7V, respectively). I am using a TM5003 mainframe to power it (I don't have a TM500). I bought this plug-in because I learned from multiple sources (including an article posted on the Vintage Tek museum website) that the TM500 plug-ins are compatible with the TM5000 mainframes (but likely not vice versa). However, after I did more testing, it seems these lower than spec voltages are due to the TM5000 supply voltages being different than the TM500 (-26V vs -33.5V and 8V vs 11.5V, respectively). Has anyone looked into this further? Can the SG503 operate at these lower voltages? It seems this would have to be the case because for example, for the -22V rail on the SG503, an LM723 regulates the incoming voltage to -22V. That's fine if it is -33V, but at -26V (actually -25V for my unit which is well within spec for TM5000) you are getting pretty close to the minimum 3V differental of the LM723.

I just want to make sure before I start going down a rabbit hole with the power supply rails.

Here's a video I did for reference on the 1st part of the SG503 repair

Thanks!

-Frank

I have a couple of the SG503's, both older and newer versions. I'll stick them in a TM500 and TM5000 mainframe and take some measurements in the next few days.

I had one that had a faulty crystal which meant the display didn't work (The crystal only runs the display, the output is generated separately with analogue RLC networks etc). I replaced it with this part and all is well again.


Also, if your unit has Texas Instruments IC sockets, you need to replace them. They are notorious for going bad. I replace them on sight with good quality circular pin types.


Jared.

I can confirm that SG503s are capable of running quite happily in TM5000
enclosures.

On Wed, Feb 14, 2024 at 11:09?AM Jared Cabot via groups.io <jaredcabot=
[email protected]> wrote:

I have a couple of the SG503's, both older and newer versions. I'll stick
them in a TM500 and TM5000 mainframe and take some measurements in the next
few days.

I had one that had a faulty crystal which meant the display didn't work
(The crystal only runs the display, the output is generated separately with
analogue RLC networks etc). I replaced it with this part and all is well
again.


Also, if your unit has Texas Instruments IC sockets, you need to replace
them. They are notorious for going bad. I replace them on sight with good
quality circular pin types.


Jared.

--
Andy

As far as I am aware ALL tm500 plugins can be used in tm5000 modules but
tm5000 and not be used in the tm 500 frames.

One not of caution there are some special tm5000 modules that only support
1 device.

Eric

IIRC the TM5000 series does not like the AM503 (?) current amplifier.? There should be a table in the archives giving compatibility.? Almost everything is happy otherwise.


The TM5000 modules do not fit into the TM500 series frames because of the extra connector for the IEEE-488 interface.

I've seen TM5000 series modules that are 1, 2 and 3 spaces wide. IIRC, there are TM5000 frames that are 1, 2, 3 and 6 spaces wide.

This follows what the TM500 frames do, with the exception that there is a TM515.? It is a 5 unit enclosure in a rugged travel case designed to allow engineers to take a custom set of instrumentation to a remote site.? IIRC it was designed to fit under a plane seat.

Harvey

I think there was some question about the AM502, I don't have one so I
never really checked.
AM503A and AM503B are explicitly compatible with TM5000 series frames.

AM503A:

AM503B:



-Eric


On Wed, Feb 14, 2024 at 10:41?AM Harvey White <madyn@...>
wrote:

IIRC the TM5000 series does not like the AM503 (?) current amplifier.
There should be a table in the archives giving compatibility. Almost
everything is happy otherwise.


The TM5000 modules do not fit into the TM500 series frames because of
the extra connector for the IEEE-488 interface.

I've seen TM5000 series modules that are 1, 2 and 3 spaces wide. IIRC,
there are TM5000 frames that are 1, 2, 3 and 6 spaces wide.

This follows what the TM500 frames do, with the exception that there is
a TM515. It is a 5 unit enclosure in a rugged travel case designed to
allow engineers to take a custom set of instrumentation to a remote
site. IIRC it was designed to fit under a plane seat.

Harvey


On 2/14/2024 1:17 PM, Eric wrote:

As far as I am aware ALL tm500 plugins can be used in tm5000 modules but
tm5000 and not be used in the tm 500 frames.

One not of caution there are some special tm5000 modules that only

support

1 device.

Eric

On Wed, Feb 14, 2024, 1:13 PM Andy Warner <andyw@...> wrote:

I can confirm that SG503s are capable of running quite happily in TM5000
enclosures.

On Wed, Feb 14, 2024 at 11:09?AM Jared Cabot via groups.io <jaredcabot=
[email protected]> wrote:

I have a couple of the SG503's, both older and newer versions. I'll

stick

them in a TM500 and TM5000 mainframe and take some measurements in the

next

few days.

I had one that had a faulty crystal which meant the display didn't work
(The crystal only runs the display, the output is generated separately

with

analogue RLC networks etc). I replaced it with this part and all is

well

again.


Also, if your unit has Texas Instruments IC sockets, you need to

replace

them. They are notorious for going bad. I replace them on sight with

good

quality circular pin types.


Jared.

--
Andy

--
--Eric
_________________________________________
Eric Garner

Thank you all for your replies! So it looks like I've got a problem with the SG503 power supply then. Because ~4.74V is probably not enough to power the counters. And the LED drivers require 5.25V per the datasheet I found (could not find a recommended minimum). Guess I need to keep looking. If anyone would be willing to take some measurements (running the SG503 with a TM5000) I would greatly appreciate it!

-Frank

The TM5000/500 mainframes have a bulk supply from which is derived 5.0 volts.? The value of that supply is not important unless it's 8.0 volts or less.? The regulators of that time used NPN transistors, which generally want to see 3.0 volts across them (7805, as an example of technology).

One thing that can pull down regulator outputs is the famous shorted tantalum capacitor.? They don't have to be shorted, just leaky as #$@!!!!.

Look at the regulator (which is on the module).? There's generally a current limiting resistor (Tek used a bit of an odd design).? Look for a series resistor of less than an ohm in the 5 volt line.? Measure the voltage drop across that and see what it works out to current.? Older designs put a base/emitter junction across that, and use that transistor to steal base current from the pass transistor.? This limits the output voltage (reduces it in this case) when more current is drawn.? Tek doesn't do this, but the idea is similar.? Too much voltage drop on this resistor shuts down the power supply.

What's the voltage drop?

Harvey

As far as I'm aware, there are only 3-module and 6-module wide TM5000 mainframes.

Many of the TM5000 modules won't fit the TM500 mainframes not only due to the GPIB connector but they also have plastic pips on the rear of the top and bottom plates that prevent insertion into TM500 mainframes.

I seem to remember there was an additional TM500 module besides the AM503 that didn't like the TM500 mainframes, maybe a Signal Generator? I can't remember off hand now....


Jared

@Harvey

Left the lab for the day, but I did check the voltage drop across that resistor R624 0.33ohm. I can't recall what it was off the top of my head, but I'll let you know tomorrow. I thoroughly tested both power supply circuits on the SG503 and if there is an issue, it must be something loading down the supplies like you suggest.

However, I injected ~5.2V into the unit with an external power supply - it did not fix the display issue. No difference whatsoever. It may be that these are the normal voltages for the unit when it is powered from a TM5000 series mainframe.

I'll be taking another look at it tomorrow! Grateful for any further suggestions as well!

-Frank

-Frank

On Thu, 15 Feb 2024, 04:52 Jared Cabot via groups.io, <jaredcabot=
[email protected]> wrote:

I seem to remember there was an additional TM500 module besides the AM503
that didn't like the TM500 mainframes, maybe a Signal Generator? I can't
remember off hand now....

I have a TM5003, with a very noisy fan, whose sole job is to house three

AM503 current probe amplifiers, and they all work absolutely fine. They're
definitely not the A or B version, just very basic (and very fixable)
analogue ones. So I assert that there's no problem with the AM503 in TM5000
mainframes.

Chris

@Harvey

Measured voltage across 0.33ohm R624 = ~0.35V. So ~1A. Seems normal.

One thing I learned reading through the circuit description is that 5V supply is referenced from the -22V rail. Would it be a good idea to inject -22V to see if this brings the 5V rail into sepc?

I was looking at the display/counter circuit more closely - everything was looking good until I got to the state generator timing relationship waveforms. They all match except pin 12 438D - there is no reset pulses. Also, pins 6 and 7 to counter U490 are low. According to the ciruit description, these should be high.

Still need to do more digging/inspecting. Thanks for everyone's help!

-Frank

Ok, Tek did a lot of "referenced" designs.? In the scopes, typically a -50 volt supply is fully adjustable, and the other supplies depend on resistor tolerances to get the right voltages. If the -22 volts is off, as you say, then everything will be off. Don't remember the design in this, but check that first.? if the -22 is off, then *it* might be loaded too much.

Looking at the schematic:

1) R680 is the current sensing resistor, so measure the drop across this.? Typically, the capacitor manufacturers told Tek that that their parts could be run quite close to the voltage limit. Say 6.3 volt capacitors on a 5.0 volt line.? They were wrong.? I'd look on the -22 volt supply for overloads pulling that down.

The voltage drop on R680 has to be less than 0.6 volts, otherwise Q690 turns on and robs base drive to the pass transistor


If 6 and 7 are high on U490, then it presets the counter to 9, and I suspect that the detect circuits cause the display to flash, indicating over range.? So those being low makes more sense in a non over range situation.

What it looks like is that the counter chain is not getting any pulses, so nothing to count.? I'd be happier fixing that 5.2 volts, since Tek ran the TTL a bit hot at 5.2 volts.? You could track the signals through from say, test point #3, but without proper voltages?? I'd not get upset on that on.

Harvey

Thanks for all that info Harvey I really appreciate it.

But in this case, I don't think the power supply is the issue. I waffled back and forth between thinking it was and it wasn't (I'd still be interested to see what someone else measures on a working SG503 powered by a TM5000). I injected both 5.2V and -22V into the unit and it did not resolve the readout issue.

Finally after probing around a bit, I think I found the problem. I measured a strange waveform on pin 13 of U435 - a SN7490. But pin 13 is a NC pin. None of the other SN7490s had this strange signal on pin 13. So I figured I would replace it but I didn't have any SN7490s on hand (I did place an order for some yesterday in case I needed one). I do however have some SN7493s. Difference is the 7490 is a decade counter and the 7493 is a 4-bit counter. Still, I figured I would pop one in to see if it made a difference. And sure enough, I started getting numbers on the frequency readout! They weren't accurate obviously, but the display was counting along with the change in frequency now!

I'll provide update once the replacement 7490s come in, but I think I found the culprit! Thanks for everyone's help!

-Frank

I received some replacement 7490s and now this SG503 is fixed! I had to really sit down and teach myself how the 7490 works. I know what counters do but never really learned how they work. I'm still learning electronics and teach myself through repair mostly. I love a repair like this one that teaches me a new concept. A bit different than your typical power supply issue - though I did spend a bit too much time on that before moving on to the actual problem.

I'd still be curious to know what voltages are typical at the -22V and 5.2V rails for this SG503 powered via a TM5000 (not TM500). Because I am still measuring ~4.7V at the 5.2V rail even with it working. If anyone would be willing to measure that and share with the group, I would appreciate it! Could save someone else from going down the same rabbit hole that I did (if in fact this is normal).

The 10Mhz setting is still a bit scratchy - it gets unleveled at this setting (likely due to bad contact). And there are a couple cosmetic things I need to finish. As well as calibration. But that is it for this guy!

Here's the vblog of the repair on my YT channel for anyone interested.


Now all I need is to fix my CG5011 and I will be able to calibrate scopes!

-Frank

On Fri, Feb 23, 2024 at 07:12 AM, Frank Mashockie wrote:

I'd still be curious to know what voltages are typical at the -22V and 5.2V
rails for this SG503 powered via a TM5000 (not TM500). Because I am still
measuring ~4.7V at the 5.2V rail even with it working. If anyone would be
willing to measure that and share with the group, I would appreciate it!
Could save someone else from going down the same rabbit hole that I did (if in
fact this is normal).

Curious whether you scoped your 5.2V rail and checked the voltage drop across
R624? My 5.2V test point is at 5.2V until things warm up for a minute or two, and
then starts wandering from 4.7-4.9 or so on my DMMs. The scope shows a square
wave on the 5V test point that I think is due to the current limiter kicking in, but
which the DMM averages out.

When the voltage drop across R624 goes below 0.6V, I'm back up to 5.2V. My
plan now is to replace the hottest ICs on the board and see if I can get the current
down closer to the level Albert Otten reported, as I can't find any shorts or bad
transistors.

Mine currently has no oscillator output unless I replace Q300 with a 2k pot, per
the Wizard Workshop suggestion, so I'm afraid I've killed Q190 or U225. I soldered
the grounded stud on the back of Q190 to the wrong point, so that's my next stop.

Congrats on fixing your SG503. This is the plugin from hell as far as I'm concerned.

Adam

One thing to note about calibrating the sg503. If you are going for the
best accuracy you will need the special cable.

In general use this is not critical but it can have a 1 to 2% effect on the
pk to pk output. So for dial accuracy this is a just be aware of it. Using
a non special cable will not effect flatness which is good.

The special cable is 50 ohms 1% if I remember the spec correctly

Eric



On Fri, Feb 23, 2024, 12:52 PM Adam R. Maxwell via groups.io <amaxwell=
[email protected]> wrote:

On Fri, Feb 23, 2024 at 07:12 AM, Frank Mashockie wrote:

I'd still be curious to know what voltages are typical at the -22V and

5.2V

rails for this SG503 powered via a TM5000 (not TM500). Because I am

still

measuring ~4.7V at the 5.2V rail even with it working. If anyone would

be

willing to measure that and share with the group, I would appreciate it!
Could save someone else from going down the same rabbit hole that I did

(if in

fact this is normal).

Curious whether you scoped your 5.2V rail and checked the voltage drop
across
R624? My 5.2V test point is at 5.2V until things warm up for a minute or
two, and
then starts wandering from 4.7-4.9 or so on my DMMs. The scope shows a
square
wave on the 5V test point that I think is due to the current limiter
kicking in, but
which the DMM averages out.

When the voltage drop across R624 goes below 0.6V, I'm back up to 5.2V. My
plan now is to replace the hottest ICs on the board and see if I can get
the current
down closer to the level Albert Otten reported, as I can't find any shorts
or bad
transistors.

Mine currently has no oscillator output unless I replace Q300 with a 2k
pot, per
the Wizard Workshop suggestion, so I'm afraid I've killed Q190 or U225. I
soldered
the grounded stud on the back of Q190 to the wrong point, so that's my
next stop.

Congrats on fixing your SG503. This is the plugin from hell as far as I'm
concerned.

Adam

Hey Frank,

The -22V supply does have an adjustment, and the 5.2V supply is derived
from the -22V supply. It turns out that -20V/-22V*5.2V is ~4.7V, so the low
5.2V supply is likely down to the low -22V supply. Did you try adjusting
the -22V supply?

Siggi

On Thu, Feb 15, 2024 at 6:53?AM Frank Mashockie <fmashockie@...>
wrote:

@Harvey

Left the lab for the day, but I did check the voltage drop across that
resistor R624 0.33ohm. I can't recall what it was off the top of my head,
but I'll let you know tomorrow. I thoroughly tested both power supply
circuits on the SG503 and if there is an issue, it must be something
loading down the supplies like you suggest.

However, I injected ~5.2V into the unit with an external power supply - it
did not fix the display issue. No difference whatsoever. It may be that
these are the normal voltages for the unit when it is powered from a TM5000
series mainframe.

I'll be taking another look at it tomorrow! Grateful for any further
suggestions as well!

-Frank

-Frank

Thanks everyone for your replies! Just to reiterate, I'm really curious about those who are powering their SG503 through the TM5000 not TM500. The reason being is that they have very different supply voltages. For the TM5000 and TM500 they supply -26V vs -33.5V and 8.2V vs 11.5V, respectively. So I've come to the conclusion that the -20V and 4.7V that I am measuring is normal when powering it with a TM5000. Especially now that I have shown it works fine. I'll look into it further if someone can prove me wrong!

@Adam

Regarding R624, I made so many measurements on the power supply circuit I forget at this point! But both the 5.2V and -22V rails are clean and smooth signals. Very little ripple. Sounds like you have your work cut out for you with your SG503!

@Eric

Thanks for the tip! I definitely don't have that cable. I'll have to look into that.

@Siggi

I have adjusted the -22V trim pot as low as it could go which is -20V. But again, this is powering the plug-in with a TM5003. After all the troubleshooting I've done, I'd be willing to bet the rails would be normal if I was using a TM500, but I don't have one. Especially when you consider the LM723. It needs a differential of 3V. And the TM5000 series has an acceptable range for the -26V rail anywhere between -23 to -27V. Mine currently outputs -25V.

-Frank

?ann lau., 24. feb. 2024 kl. 00:46 skrifa?i Frank Mashockie <
fmashockie@...>:

@Siggi

I have adjusted the -22V trim pot as low as it could go which is -20V.
But again, this is powering the plug-in with a TM5003. After all the
troubleshooting I've done, I'd be willing to bet the rails would be normal
if I was using a TM500, but I don't have one. Especially when you consider
the LM723. It needs a differential of 3V. And the TM5000 series has an
acceptable range for the -26V rail anywhere between -23 to -27V. Mine
currently outputs -25V.

Hey Frank,

It's true that the LM723 needs ~3V of headroom in a series configuration,
but that's not how it's being used in the -22V supply. Here it's in more of
a shunt configuration, and what determines the dropout voltage of the
circuit here (AFAICT) is the Vbe of Q685, the Vbe of the mainframe pass
transistor and the drop over R680. Depending on the expected current drawn
from this supply, I think -25V should give it plenty of headroom. A
ballpark is 3*0.7V-25V=-22.9V, meaning there should be 0.9V of headroom
remaining at the point where the current limit trips.
Maybe your =22V supply is overloaded - note that if the voltage across R680
exceeds 0.6-0.7V, the rail will current limit.

Also, If you measure the voltage across pins 4 and 5 of the LM723, you
should be able to see whether it's in regulation or not. If that voltage is
not as close to zero as makes no difference, your LM723 is not in
regulation. If it's in regulation, the problem might be due to drift in the
sense or reference resistor networks.

Good luck,
Siggi