UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Test Gear and Workshop Equipment

Notices

Vintage Test Gear and Workshop Equipment For discussions about vintage test gear and workshop equipment such as coil winders.

Closed Thread
 
Thread Tools
Old 6th May 2020, 12:41 pm   #61
allan
Pentode
 
allan's Avatar
 
Join Date: Oct 2005
Location: South coast near Ringwood/Christchurch, Dorset, UK.
Posts: 230
Default Re: HP8640B Signal Generator

I've ordered a set of chrome vanadium Allen Keys from RS (151-1519) so the 0.05 and sixteenth etc should be here tomorrow. I looked for ages last week and just discovered the 0.05 is wrongly listed as 0.5 inch. Let's hope the screws aren't chewed up.
I'm hot on the trail of the meter reading problem. Working backwards from the meter.
The signal crosses 2 or 3 motherboards and risers then disappears into a mode switch that probably outputs a binary code. The earlier circuit I have seems to indicate that previous models used a resistor chain. If I'm right the engineers must have been let loose to redesign things. It was probably the start of the digital revolution...literally.
The mode switch does appear to have intact wipers so I'll be looking for a decoder next, but that of course will not be in the older manual so I'll need to switch to a later version. Slightly worrying is the small pot close to the mode switch has been messily resoldered so I'm probably not the first here.
I'll be going quiet for a time later as there's a lift repair due in today or tomorrow. An obsolete circuit board but likely not as obsolete as the HP8640B...
Allan
Attached Thumbnails
Click image for larger version

Name:	attenuator coder.jpg
Views:	107
Size:	60.2 KB
ID:	205059  
allan is offline  
Old 6th May 2020, 12:51 pm   #62
Andrewausfa
Octode
 
Andrewausfa's Avatar
 
Join Date: Dec 2010
Location: St Ives, Cambridgeshire, UK.
Posts: 1,180
Default Re: HP8640B Signal Generator

Having seen the state those pair were in Allan, you're doing a good job at getting it up and running again. Enjoyed reading through your tribulations so far.

Andrew
__________________
Invisible airwaves crackle with life. Or they should do.
BVWS Member
Andrewausfa is offline  
Old 6th May 2020, 1:36 pm   #63
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,799
Default Re: HP8640B Signal Generator

If you need any guidance on that attenuator, look up "Art Fong Attenuator" there's an article in, I think the HP memory project.

Art Fong himself is well worth looking up, too.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 6th May 2020, 2:47 pm   #64
allan
Pentode
 
allan's Avatar
 
Join Date: Oct 2005
Location: South coast near Ringwood/Christchurch, Dorset, UK.
Posts: 230
Default Re: HP8640B Signal Generator

I wasn't casting nasturtiums on Mr Fong's attenuator, just engineers in general David.
http://www.radiomuseum.co.uk/lamp.html
You can guess one of my pet hates is overcomplication ever since I repaired a chip pan using a microprocessor and Mr Dysons immortal vaccuum cleaners with broken mains leads.
I've now discovered why the attenuator has a big gearwheel mounted behind the knob. I was panicking thinking someone had removed something really vital until I spotted pictures of a version fitted with a long geared axle (maybe a fine tune for the attenuator?)
I also found a high def picture on that memory project from which I might be able to make a new attenuator dial. The one on mine has loads of corrosion which I haven't tried to clean off yet.
I haven't decided about supergluing/replacing the various gears yet. If I can remove the remaining knobs I might extract the assemblies and remove the gears so I don't get superglue on the axles.
I also have some "milliput" epoxy putty which might work better?
Most of the plastic gears have cracks, some just in their hub. In the past, for other equipments, I've glued a metal washer to the back to hold them together.
Then I've got to inspect the mode switches on the range and peak deviation switches because I don't think the range changes.
Allan
allan is offline  
Old 9th May 2020, 2:02 pm   #65
Nymrod121
Nonode
 
Nymrod121's Avatar
 
Join Date: Apr 2012
Location: Worcestershire, UK.
Posts: 2,039
Default Re: HP8640B Signal Generator

The attached may help towards further cosmetic refurbishment
I scanned in the front panel of my ‘spares mule’ ‘8640B – it had to be done as two separate images as my scanner is A4 (297mm x 210mm) and the panel dimensions are 400 x 76mm. I then set to with Microsoft Paint for a jolly hour or three and edited out as many of the scuffs, blemishes etc. as possible.
I think the end result is pretty good ... but as to how much it would cost to have one made up for real is anyone’s guess.
All the best
Guy
Attached Files
File Type: pdf HP8640B_front panel as scanned.pdf (247.5 KB, 71 views)
File Type: pdf HP8640B front panel_v01c.pdf (294.2 KB, 63 views)
__________________
"What a depressingly stupid machine." [Marvin: HHGTTG]
Nymrod121 is offline  
Old 9th May 2020, 2:08 pm   #66
Nymrod121
Nonode
 
Nymrod121's Avatar
 
Join Date: Apr 2012
Location: Worcestershire, UK.
Posts: 2,039
Default Re: HP8640B Signal Generator

I went through a similar exercise with the 'reverse power protection' label which was originally fixed over the 'Do Not Apply Reverse Power' legend on the front panel
All the best
Guy
__________________
"What a depressingly stupid machine." [Marvin: HHGTTG]
Nymrod121 is offline  
Old 9th May 2020, 3:36 pm   #67
allan
Pentode
 
allan's Avatar
 
Join Date: Oct 2005
Location: South coast near Ringwood/Christchurch, Dorset, UK.
Posts: 230
Default Re: HP8640B Signal Generator

Very much a Work-in-Progress but I'm slowly getting a working HP8640B.
Being wise after the event is very true as I fault-find particularly noticing whether a meter reading is plus or minus. With a modern multimeter not always obvious.
Next, before aligning the AGC system, I'm looking at the three lamps for advising on which scale is relevant. No doubt an LED solution...
Progress here (but not checked for typos etc)
http://www.radiomuseum.co.uk/HP8640B.html
Allan
PS How do I get at the lamps on the RHS of the meter!!

Last edited by allan; 9th May 2020 at 3:42 pm.
allan is offline  
Old 9th May 2020, 5:51 pm   #68
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,799
Default Re: HP8640B Signal Generator

Superglue will fix the cracks in the gears, only for them to crack somewhere else. I'm afraid the whole things have turned brittle with time.

I suppose it's glue for the time being, but once you get the thing going and decide it's a goer, then the outlay on those brass parts from India may be worthwhile.

The 8640 isn't just a comprehensive signal generator, it is also particularly low noise and can form part of a phase noise test setup.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 10th May 2020, 10:39 am   #69
allan
Pentode
 
allan's Avatar
 
Join Date: Oct 2005
Location: South coast near Ringwood/Christchurch, Dorset, UK.
Posts: 230
Default Re: HP8640B Signal Generator

I finally diagnosed the final fault in my HP8640B that resulted in it becoming a "barn find". The detector diode and it's series resistor had been zapped by a transceiver blipping into transmit.
My quick fix is a simple diode detector external to the RF output hybrid which produces a feedback voltage (and a meter voltage) at 4MHz.
Looking at the original circuit I guess it's much less frequency dependent than mine which links the 20dBm RF output via a 33pF capacitor.
The original circuit seems to monitor the current through a 35 ohm resistor.

I wonder if anyone has tackled a fault in the detector?

A better solution might be to open the output and insert a small resistor then link this via a 0.1uF chip to the coax. The quick fix works but the 33pF capacitor will mean the detected output voltage will drop off as the frequency decreases.
I could increase this to say 0.1uF but the original circuit seems to monitor the RF current through R1, 35 ohms (and the 50 ohm attenuator?) and apart from a smoothing capacitor C3 will be much less frequency dependent.
I wonder if anyone has any suggestions before I start further experiments?
Incidentally R22 and R23 provide the forward bias of circa +0.6 volts for the detector diode and R3, 1K is the path to ground for this.


Allan
Attached Thumbnails
Click image for larger version

Name:	detector.jpg
Views:	77
Size:	30.4 KB
ID:	205392  

Last edited by allan; 10th May 2020 at 10:52 am.
allan is offline  
Old 10th May 2020, 2:55 pm   #70
Nymrod121
Nonode
 
Nymrod121's Avatar
 
Join Date: Apr 2012
Location: Worcestershire, UK.
Posts: 2,039
Default Re: HP8640B Signal Generator

Quote:
Originally Posted by allan View Post
Very much a Work-in-Progress but I'm slowly getting a working HP8640B.
Being wise after the event is very true as I fault-find particularly noticing whether a meter reading is plus or minus. With a modern multimeter not always obvious.
Next, before aligning the AGC system, I'm looking at the three lamps for advising on which scale is relevant. No doubt an LED solution...
Progress here (but not checked for typos etc)
http://www.radiomuseum.co.uk/HP8640B.html
Allan
PS How do I get at the lamps on the RHS of the meter!!
They're integrated within the frequency counter module and are mounted within plastic mouldings such that each bulb is soldered to the board with the leads turned through 90 degrees to give the appropriate alignment for presentation forwards through the front panel. For obvious reasons, those were the only two that I didn't replace on my '86040B when I did an "LED conversion" job on the meter annunciator PCB assembly.

I'll have a a look for my spares mule counter, take the lid off and post a pic ... watch this space

Guy
__________________
"What a depressingly stupid machine." [Marvin: HHGTTG]
Nymrod121 is offline  
Old 10th May 2020, 2:58 pm   #71
factory
Octode
 
Join Date: Jul 2009
Location: Cheshire, UK.
Posts: 1,706
Default Re: HP8640B Signal Generator

I'm not sure what you could do inside the hybrid, someone did repair one with a vaporised bond wire (scroll down to 14.);
https://www.ve7ca.net/TstH86.htm#HK2

Also they have a cheaper way of repairing the gears, you may have already seen this website as the hybrid circuit diagram is on there.

David
factory is online now  
Old 10th May 2020, 3:34 pm   #72
Nymrod121
Nonode
 
Nymrod121's Avatar
 
Join Date: Apr 2012
Location: Worcestershire, UK.
Posts: 2,039
Default Re: HP8640B Signal Generator

Quote:
Originally Posted by Nymrod121 View Post
They're integrated within the frequency counter module and are mounted within plastic mouldings such that each bulb is soldered to the board with the leads turned through 90 degrees to give the appropriate alignment for presentation forwards through the front panel. For obvious reasons, those were the only two that I didn't replace on my '86040B when I did an "LED conversion" job on the meter annunciator PCB assembly.

I'll have a a look for my spares mule counter, take the lid off and post a pic ... watch this space

Guy
Here you go ...
(1st lot of five)
Attached Thumbnails
Click image for larger version

Name:	20200510_150350.jpg
Views:	73
Size:	51.0 KB
ID:	205421   Click image for larger version

Name:	20200510_150432.jpg
Views:	84
Size:	66.9 KB
ID:	205422   Click image for larger version

Name:	20200510_150654.jpg
Views:	86
Size:	69.5 KB
ID:	205423   Click image for larger version

Name:	20200510_150704.jpg
Views:	76
Size:	51.4 KB
ID:	205424   Click image for larger version

Name:	20200510_151111.jpg
Views:	79
Size:	120.6 KB
ID:	205425  

__________________
"What a depressingly stupid machine." [Marvin: HHGTTG]
Nymrod121 is offline  
Old 10th May 2020, 3:35 pm   #73
Nymrod121
Nonode
 
Nymrod121's Avatar
 
Join Date: Apr 2012
Location: Worcestershire, UK.
Posts: 2,039
Default Re: HP8640B Signal Generator

2nd lot of three

(darn it ... 'Operator J' interfered with the last one )
Attached Thumbnails
Click image for larger version

Name:	20200510_151131.jpg
Views:	68
Size:	125.6 KB
ID:	205426   Click image for larger version

Name:	20200510_151139.jpg
Views:	88
Size:	71.4 KB
ID:	205427   Click image for larger version

Name:	20200510_151151.jpg
Views:	72
Size:	73.7 KB
ID:	205428  
__________________
"What a depressingly stupid machine." [Marvin: HHGTTG]
Nymrod121 is offline  
Old 10th May 2020, 3:44 pm   #74
allan
Pentode
 
allan's Avatar
 
Join Date: Oct 2005
Location: South coast near Ringwood/Christchurch, Dorset, UK.
Posts: 230
Default Re: HP8640B Signal Generator

Lamps all seem to be OK as their filaments are intact folks.
The fault seems to be the ODD/EVEN switch which selects either 0-10 or 0-3 meter scales.
I'm now tracking down that fault....
Allan
allan is offline  
Old 10th May 2020, 6:01 pm   #75
IanG3XYV
Pentode
 
Join Date: Sep 2014
Location: Bedford, Bedfordshire, UK.
Posts: 172
Default Re: HP8640B Signal Generator

Catching up on this fascinating thread reminds me that I really need to get my two 8640Bs out to check (when I clear some bench space that is). I think they were fine when last used a year or two ago but due an annual service I think.

As you'll have found there is lots of advice on the web, often to do with the rotary switches and gears and one of the best sites that I have just found is here:
https://k6jca.blogspot.com/search/label/HP%208640B

Ian
IanG3XYV is offline  
Old 10th May 2020, 6:17 pm   #76
Nymrod121
Nonode
 
Nymrod121's Avatar
 
Join Date: Apr 2012
Location: Worcestershire, UK.
Posts: 2,039
Default Re: HP8640B Signal Generator

Quote:
Originally Posted by IanG3XYV View Post
(snipped)

As you'll have found there is lots of advice on the web, often to do with the rotary switches and gears and one of the best sites that I have just found is here:
https://k6jca.blogspot.com/search/label/HP%208640B

Ian
Thanks for ths, Ian - it is certainly the best one I've seen in terms of a clear, concise step-by-step description of servicing the A9 assembly (and K6JCA's example uses the clear plastic rotors, same as Allan's ... and also my 'workshop' one - whereas my 'spares mule' is a non-doubler version equipped with the earlier wafer switch assembly)


Guy
__________________
"What a depressingly stupid machine." [Marvin: HHGTTG]
Nymrod121 is offline  
Old 10th May 2020, 11:03 pm   #77
allan
Pentode
 
allan's Avatar
 
Join Date: Oct 2005
Location: South coast near Ringwood/Christchurch, Dorset, UK.
Posts: 230
Default Re: HP8640B Signal Generator

I inspected the mode switches on the geared assembly a week ago and all the wipers looked OK so breathed a sigh of relief. Just the gears to fix.
Short-lived however because the rotary switch on the back of the attenuator was where I ended up looking for the missing ODD signal that flips the 0-3/0-10 lamps and I found the end wiper was missing.
I managed to detach the clear plastic rotor and found also the centre wiper was hanging on by only one pip.
I looked everywhere for a new wiper and eventually broke open an edge connector (I have a box of these green things that are the same type as used in the 8640B) The gold flashed springs are the right width so after a bit of cut and pasting the new one, that joins the two originals and all three are now superglued in place with a 2-part epoxy to complete the job. Tomorrow I'll attempt to reassemble it.

You'll need a spot of spinach for breakfast Ian if you intend to give your sig gen the once over.

I'd missed that excellent article by K6JCA.
Allan
allan is offline  
Old 11th May 2020, 7:56 pm   #78
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: HP8640B Signal Generator

Quote:
Originally Posted by allan View Post
I finally diagnosed the final fault in my HP8640B that resulted in it becoming a "barn find". The detector diode and it's series resistor had been zapped by a transceiver blipping into transmit.
My quick fix is a simple diode detector external to the RF output hybrid which produces a feedback voltage (and a meter voltage) at 4MHz.
Looking at the original circuit I guess it's much less frequency dependent than mine which links the 20dBm RF output via a 33pF capacitor.
The original circuit seems to monitor the current through a 35 ohm resistor.

I wonder if anyone has tackled a fault in the detector?

A better solution might be to open the output and insert a small resistor then link this via a 0.1uF chip to the coax. The quick fix works but the 33pF capacitor will mean the detected output voltage will drop off as the frequency decreases.
I could increase this to say 0.1uF but the original circuit seems to monitor the RF current through R1, 35 ohms (and the 50 ohm attenuator?) and apart from a smoothing capacitor C3 will be much less frequency dependent.
I wonder if anyone has any suggestions before I start further experiments?
Incidentally R22 and R23 provide the forward bias of circa +0.6 volts for the detector diode and R3, 1K is the path to ground for this.


Allan
I'm not that familiar with the HP8640 design but most signal generators would have a value of 50 ohm for R1 in your diagram below. Normally, this series resistor sets the source impedance for the sig gen.
This is the most significant and important when the attenuator is set to 0dB as R1 is the only resistive element at a 0dB setting.

Therefore, it isn't a good idea to put the ALC diode after R1 (35 ohm) as the generator source impedance will be reduced a lot when 0dB is selected. In theory at least, the circuit node where the ALC diode sits will act as a low Z voltage source when the ALC feedback/correction loop is closed and working properly. You can think of the ALC as a system that maintains an extremely 'stiff' RF voltage at the C1 and R1 node. The stiffness is reinforced by the feedback correction so this node will be like a voltage source. It's the RF equivalent of achieving an ultra low source impedance from a dc voltage regulator chip. Within the bandwidth of the correction loop the feedback creates a low Z voltage source.

So most sig gens follow this 'stiff' circuit node at C1/R1 with a series 50R resistor at R1 to help preserve the 50 ohm source impedance at low attenuation settings. HP used 35R here and there must be a valid reason for this. Maybe a compromise for best source impedance at 0dB across 0.5-1024MHz? There may be some additional ESR in the attenuator switching elements for example and this may get worse with increasing frequency. So setting R1 at 35R may be better than 50R here because of the added series resistance in the attenuator at the 0dB setting. But I'm really just guessing.
Attached Thumbnails
Click image for larger version

Name:	8640detector.jpg
Views:	62
Size:	30.4 KB
ID:	205538  
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 11th May 2020 at 8:12 pm.
G0HZU_JMR is offline  
Old 11th May 2020, 8:35 pm   #79
factory
Octode
 
Join Date: Jul 2009
Location: Cheshire, UK.
Posts: 1,706
Default Re: HP8640B Signal Generator

Quote:
Originally Posted by G0HZU_JMR View Post
I'm not that familiar with the HP8640 design but most signal generators would have a value of 50 ohm for R1 in your diagram below. Normally, this series resistor sets the source impedance for the sig gen.
This is the most significant and important when the attenuator is set to 0dB as R1 is the only resistive element at a 0dB setting.

Therefore, it isn't a good idea to put the ALC diode after R1 (35 ohm) as the generator source impedance will be reduced a lot when 0dB is selected.
Unfortunately all the parts from the original circuit are inside the 08640-67002 hybrid IC, unless the top is cut off that & micro-surgery attempted it isn't possible to put the detector diode in the correct place.

Or there are untested pulls available, what is the chance they will also be the reason the 8640 was parted out by the sellers of these?

David
factory is online now  
Old 12th May 2020, 9:08 am   #80
allan
Pentode
 
allan's Avatar
 
Join Date: Oct 2005
Location: South coast near Ringwood/Christchurch, Dorset, UK.
Posts: 230
Default Re: HP8640B Signal Generator

Going back to the output monitor diode.
I could eventually cut the top off the hybrid and see if I can connect a diode or even see if its just the series resistor that failed if all else fails.
I can't remember exactly, but there are a couple of resistors in the monitoring system that may be used to set the AGC feedback parameters, of which one is used solely for +20dBm (no attenuation) and the other for all attenuated settings.
Yes, those untested pulls might well be duff David.

I'm still struggling. The attenuator knob refuses to part with its shaft but I may not need it to come off because I'm just finishing repair of the rear mounted switch with the attenuator in-situ. Of course if the latter was damaged I'll have to cut off the knob to get the thing out and the lid off etc.

A wobbly coax feed to the oscillator fell off yesterday because the bulkhead socket thread had sheared (before I got it). Hopefully I can remove the oscillator assembly to fix (a) the broken output socket (b) the disks the limit the tuning range and (c) glue cracked gearwheels.
Is this oscillator assembly easy to remove??
Allan
Attached Thumbnails
Click image for larger version

Name:	osc.jpg
Views:	84
Size:	64.9 KB
ID:	205570  
allan is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 3:21 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.