Advance E2 Signal Generator fault - UK Vintage Radio Repair and Restoration Discussion Forum

Andrew Wyard · 16th Jun 2021, 6:31 pm

Hi Chaps hope you can help.

I have an E2 sig gen bought from a rally some time back.
From the start it had issues, most of which haven't been that much improved.
The lower three frequency ranges (D,E,F) are pretty much as they should be with good amplitude. The upper three get successively worse with frequency.
Across the range- F=2.5V E=2V, D=1V to 2V, C=1V, B=0.6V to 0.08V, A staggers in to life mid range at about 0.1V All measured at full RF OP and also on tag strip.

Initially it had low HT, so replacement 6X5 and filter caps that helped that. Voltages quoted on diagram pretty much match.
I also ordered a new ECC91 this has possibly helped a little.
I have thoroughly cleaned the wafer switch.
I have replaced the C and R connected to the osc grid.
I have checked the VC for leakage.
I have tried disconnecting the attenuator stage to see if that was loading it.
I have also checked and changed some other resistors around the circuit that weren't that close to value

My last thought is the possibility that the ECC91 is the problem, even though it and the original perform in a similar way.

Any way that's probably enough waffle for now, hopefully there's somebody that's gone through some of this before and has some ideas.

Many thanks, Andrew Wyard. PS if I don't come back to responses for a bit, its just work etc getting in the way. Thanks.

Electronpusher0 · 17th Jun 2021, 8:53 am

Hello Andrew and welcome to the forum.

You seem to have tried most things but have you tried disconnecting the full output socket?
The E2 has an additional socket on the front panel that provides the full RF output, this was supplied with a rotating shutter that could be slid round to block off the socket to prevent accidental use.
If this socket went faulty in some way it could drag down the output especially at higher frequencies.
Just a thought.

Peter

Andrew Wyard · 17th Jun 2021, 9:38 am

Hi Peter,
Yes, I disconnected the feed to all the outputs on the tag strip under the wafer sw.

I think over the weekend I will have another look at readings across the op coupling windings, I think they were fairly sensible but there's something pulling it down. I suppose its possible someone has put some volts on the full RF connector and partly fried one or more op winding.

Thanks, Andrew

orbanp1 · 17th Jun 2021, 12:31 pm

Hi Andrew,

Just curious, how did you measure the output voltages?
Any chance that the low readings are the results of the measurements?

Regards, Peter

MotorBikeLes · 17th Jun 2021, 10:23 pm

I had quite a few E2s 40 years ago, but I never saw one with an ECC91 as best as I can recall. I think all mine had octal valves, and only one half of the valve was used. I once had one with intermittent output, but that was a badly (read NEVER) soldered connection to the valve base.
I would be inclined to try a different valve. I don't know the characteristics of the ECC91, but maybe half an ECC88 would be better. Why not a single valve instead of a double. Would an EC92 be suitable. Just random ideas.
The output DOES fall on higher ranges, but that seems rather a lot.
Les.

Chris55000 · 17th Jun 2021, 11:16 pm

Hi!

The booklet for the E2 on "Jon's Valve Page" looks to be a more modern booklet, that does make reference to an ECC91/6J6 valve in the RF oscillator box – the diagram incorrectly labels it as "ECC82!"

The ECC91 can be found all over the 1968 edition of the RSGB "Radio Communication Handbook", it was used for virtually almost anything!

It has an anode resistance of 7100 ohms and a gm of 5.5 mA/V at –1.5V GB.

This valve is plentiful and cheap n.o.s., typically £5 plus £3 p and p, so I wouldn't bother guessing about its condition or substituting it!

Chris Williams

ms660 · 18th Jun 2021, 10:51 am

Quote:

Originally Posted by Chris55000

The booklet for the E2 on "Jon's Valve Page" looks to be a more modern booklet, that does make reference to an ECC91/6J6 valve in the RF oscillator box – the diagram incorrectly labels it as "ECC82!"

That's a re-drawn job, amongst other things V2's heater pin numbers are incorrect.

Electrotanya have the manual.

Lawrence.

Andrew Wyard · 21st Jun 2021, 9:18 pm

Hi, thanks for the reply's, not had much time to get back to this unfortunately.
To reply to Peter-old Tek oscilloscope for measurements.
I will obtain another oscillator valve, also the EC91 looks possible too.
I do have the re drawn circuit but might try to get the proper manual.

will let you know how it goes. Thanks , Andrew

WME_bill · 21st Jun 2021, 9:46 pm

Advance E2.
The Volkswagen of servicing signal generators. I have looked at mine, and indeed the output does reduce a bit on the highest range. But if you look at the coil at the 30 to 100 Mhz range, you will see it is just the strap from the switch to the very large variable capacitor. The Q of that circuit must be minimal, so any trace of dirt or grease could easily be enough to dampen the oscillation. If you look at the scale, it just about does down to 30Mhz and there is a long gap to the end of the low frequency end of the range.
The amplitude of oscillation increases as you go towards 100Mhz (Range A).
I had to try several 6J6 to get the best one, selecting from my stock of NOS and used ones. Clean the valve pins and base.

Valve types. The ECC91 or 6J6 was one of the better valves for oscillator use. The ECC81 / 12AT7 was slightly better, as you see it used often in more modern signal generators.
A figure of merit for an oscillator valve can be otained from gmXgm X Ra. (Langford Smith Radio Designer Handbook 4th edn, chapter 24.2, page 950) .
The ECC82 / 12AU7 / 6SN7 /6J5 is less effective. Used in the Advance E1 and the Avo Sig Gen, covering only up to about 50 Mhz.

Merit: 6J6= 200, ECC81 = 385, ECC82 = 60, ECC88 = 405. The star performer used in Taylor 67 is EC91 = 870,but a rare valve. And the TV UHF oscillator types PC86, PC97 = 850-900.
Or for FET, the 2N3821 / 2N5457.

I have used a EF91 triode connected = 415 in a grid dip oscillator when a 6C4 =35 did not go high enough, after a 9002 (Acorn 955 on B7G base) failed.
wme_bill

DangerMan · 21st Jun 2021, 11:00 pm

Hi Andrew,

What is the bandwidth of your oscilloscope, and what cable/probe type are you using?

Not trying to teach you how to suck eggs but a lack of bandwidth here could explain the symptom at high frequencies.

Pete

Chris55000 · 22nd Jun 2021, 10:58 am

Hi!

Without going to "valvephoolery" prices, the OP can get a reasonably priced EC91/CV417/6AQ4 from this link:–

https://www.ebay.co.uk/itm/Mullard-K....m46890.l49286

The EC91/CV417 should work if the leads to pins 5 and 7 are swapped over, and any wiring to pins 1 or 2 transferred to pin 7, and anything on pin 5 transferred to pin 6!

ECC91 (original) base:–

1:– a', 2:– a", 3:– h, 4:– h, 5:– g' 6:– g", 7:- k

EC91/CV417 (suggested):–

1:– k, 2:– g, 3:– h, 4:– h, 5:– k, 6:– g, 7:– a

Chris Williams

Andrew Wyard · 23rd Jun 2021, 3:25 pm

Hi, just a quick reply.

I'm not going to get near this again until next week!

As for the test equipment. It's a Tektronix 454 which is quoted on certain ranges to go up to 150MHz. The probe might be a different story. It's not an original Tek item, I will do some checking to make sure I'm not being taken for a ride. I was using another probe which gave similar results, but it would be good to rule them out.

Thanks for the comments. Andrew.

16th Jun 2021, 6:31 pm	#1
Andrew Wyard Diode Join Date: Mar 2021 Location: Bracknell, Berkshire, UK Posts: 4	Advance E2 Signal Generator fault Hi Chaps hope you can help. I have an E2 sig gen bought from a rally some time back. From the start it had issues, most of which haven't been that much improved. The lower three frequency ranges (D,E,F) are pretty much as they should be with good amplitude. The upper three get successively worse with frequency. Across the range- F=2.5V E=2V, D=1V to 2V, C=1V, B=0.6V to 0.08V, A staggers in to life mid range at about 0.1V All measured at full RF OP and also on tag strip. Initially it had low HT, so replacement 6X5 and filter caps that helped that. Voltages quoted on diagram pretty much match. I also ordered a new ECC91 this has possibly helped a little. I have thoroughly cleaned the wafer switch. I have replaced the C and R connected to the osc grid. I have checked the VC for leakage. I have tried disconnecting the attenuator stage to see if that was loading it. I have also checked and changed some other resistors around the circuit that weren't that close to value My last thought is the possibility that the ECC91 is the problem, even though it and the original perform in a similar way. Any way that's probably enough waffle for now, hopefully there's somebody that's gone through some of this before and has some ideas. Many thanks, Andrew Wyard. PS if I don't come back to responses for a bit, its just work etc getting in the way. Thanks.

17th Jun 2021, 8:53 am	#2
Electronpusher0 Nonode Join Date: Jun 2016 Location: Bognor Regis, West Sussex, UK. Posts: 2,294	Re: Advance E2 Signal Generator fault Hello Andrew and welcome to the forum. You seem to have tried most things but have you tried disconnecting the full output socket? The E2 has an additional socket on the front panel that provides the full RF output, this was supplied with a rotating shutter that could be slid round to block off the socket to prevent accidental use. If this socket went faulty in some way it could drag down the output especially at higher frequencies. Just a thought. Peter

17th Jun 2021, 9:38 am	#3
Andrew Wyard Diode Join Date: Mar 2021 Location: Bracknell, Berkshire, UK Posts: 4	Re: Advance E2 Signal Generator fault Hi Peter, Yes, I disconnected the feed to all the outputs on the tag strip under the wafer sw. I think over the weekend I will have another look at readings across the op coupling windings, I think they were fairly sensible but there's something pulling it down. I suppose its possible someone has put some volts on the full RF connector and partly fried one or more op winding. Thanks, Andrew

17th Jun 2021, 12:31 pm	#4
orbanp1 Heptode Join Date: Nov 2009 Location: London, Ontario, Canada Posts: 675	Re: Advance E2 Signal Generator fault Hi Andrew, Just curious, how did you measure the output voltages? Any chance that the low readings are the results of the measurements? Regards, Peter

17th Jun 2021, 10:23 pm	#5
MotorBikeLes Nonode Join Date: Dec 2012 Location: Kirk Michael, Isle of Man Posts: 2,350	Re: Advance E2 Signal Generator fault I had quite a few E2s 40 years ago, but I never saw one with an ECC91 as best as I can recall. I think all mine had octal valves, and only one half of the valve was used. I once had one with intermittent output, but that was a badly (read NEVER) soldered connection to the valve base. I would be inclined to try a different valve. I don't know the characteristics of the ECC91, but maybe half an ECC88 would be better. Why not a single valve instead of a double. Would an EC92 be suitable. Just random ideas. The output DOES fall on higher ranges, but that seems rather a lot. Les.

17th Jun 2021, 11:16 pm	#6
Chris55000 Nonode Join Date: Aug 2007 Location: Walsall Wood, Aldridge, Walsall, UK. Posts: 2,864	Re: Advance E2 Signal Generator fault Hi! The booklet for the E2 on "Jon's Valve Page" looks to be a more modern booklet, that does make reference to an ECC91/6J6 valve in the RF oscillator box – the diagram incorrectly labels it as "ECC82!" The ECC91 can be found all over the 1968 edition of the RSGB "Radio Communication Handbook", it was used for virtually almost anything! It has an anode resistance of 7100 ohms and a gm of 5.5 mA/V at –1.5V GB. This valve is plentiful and cheap n.o.s., typically £5 plus £3 p and p, so I wouldn't bother guessing about its condition or substituting it! Chris Williams __________________ It's an enigma, that's what it is! This thing's not fixed because it doesn't want to be fixed!

21st Jun 2021, 9:18 pm	#8
Andrew Wyard Diode Join Date: Mar 2021 Location: Bracknell, Berkshire, UK Posts: 4	Re: Advance E2 Signal Generator fault Hi, thanks for the reply's, not had much time to get back to this unfortunately. To reply to Peter-old Tek oscilloscope for measurements. I will obtain another oscillator valve, also the EC91 looks possible too. I do have the re drawn circuit but might try to get the proper manual. will let you know how it goes. Thanks , Andrew

21st Jun 2021, 9:46 pm	#9
WME_bill Octode Join Date: Oct 2007 Location: Exeter, Devon, UK. Posts: 1,554	Re: Advance E2 Signal Generator fault Advance E2. The Volkswagen of servicing signal generators. I have looked at mine, and indeed the output does reduce a bit on the highest range. But if you look at the coil at the 30 to 100 Mhz range, you will see it is just the strap from the switch to the very large variable capacitor. The Q of that circuit must be minimal, so any trace of dirt or grease could easily be enough to dampen the oscillation. If you look at the scale, it just about does down to 30Mhz and there is a long gap to the end of the low frequency end of the range. The amplitude of oscillation increases as you go towards 100Mhz (Range A). I had to try several 6J6 to get the best one, selecting from my stock of NOS and used ones. Clean the valve pins and base. Valve types. The ECC91 or 6J6 was one of the better valves for oscillator use. The ECC81 / 12AT7 was slightly better, as you see it used often in more modern signal generators. A figure of merit for an oscillator valve can be otained from gmXgm X Ra. (Langford Smith Radio Designer Handbook 4th edn, chapter 24.2, page 950) . The ECC82 / 12AU7 / 6SN7 /6J5 is less effective. Used in the Advance E1 and the Avo Sig Gen, covering only up to about 50 Mhz. Merit: 6J6= 200, ECC81 = 385, ECC82 = 60, ECC88 = 405. The star performer used in Taylor 67 is EC91 = 870,but a rare valve. And the TV UHF oscillator types PC86, PC97 = 850-900. Or for FET, the 2N3821 / 2N5457. I have used a EF91 triode connected = 415 in a grid dip oscillator when a 6C4 =35 did not go high enough, after a 9002 (Acorn 955 on B7G base) failed. wme_bill Last edited by WME_bill; 21st Jun 2021 at 9:53 pm.

21st Jun 2021, 11:00 pm	#10
DangerMan Heptode Join Date: Jul 2009 Location: Stoke-on-Trent, Staffordshire, UK. Posts: 719	Re: Advance E2 Signal Generator fault Hi Andrew, What is the bandwidth of your oscilloscope, and what cable/probe type are you using? Not trying to teach you how to suck eggs but a lack of bandwidth here could explain the symptom at high frequencies. Pete

22nd Jun 2021, 10:58 am	#11
Chris55000 Nonode Join Date: Aug 2007 Location: Walsall Wood, Aldridge, Walsall, UK. Posts: 2,864	Re: Advance E2 Signal Generator fault Hi! Without going to "valvephoolery" prices, the OP can get a reasonably priced EC91/CV417/6AQ4 from this link:– https://www.ebay.co.uk/itm/Mullard-K....m46890.l49286 The EC91/CV417 should work if the leads to pins 5 and 7 are swapped over, and any wiring to pins 1 or 2 transferred to pin 7, and anything on pin 5 transferred to pin 6! ECC91 (original) base:– 1:– a', 2:– a", 3:– h, 4:– h, 5:– g' 6:– g", 7:- k EC91/CV417 (suggested):– 1:– k, 2:– g, 3:– h, 4:– h, 5:– k, 6:– g, 7:– a Chris Williams __________________ It's an enigma, that's what it is! This thing's not fixed because it doesn't want to be fixed! Last edited by Chris55000; 22nd Jun 2021 at 11:20 am.

23rd Jun 2021, 3:25 pm	#12
Andrew Wyard Diode Join Date: Mar 2021 Location: Bracknell, Berkshire, UK Posts: 4	Re: Advance E2 Signal Generator fault Hi, just a quick reply. I'm not going to get near this again until next week! As for the test equipment. It's a Tektronix 454 which is quoted on certain ranges to go up to 150MHz. The probe might be a different story. It's not an original Tek item, I will do some checking to make sure I'm not being taken for a ride. I was using another probe which gave similar results, but it would be good to rule them out. Thanks for the comments. Andrew.