How could I tell? HP8601A RF Sweep Generator. - UK Vintage Radio Repair and Restoration Discussion Forum

georgesgiralt · 10th Dec 2009, 3:37 pm

Hi !
Yesterday, I was playing around with my "brand new" HP8601A RF sweep generator (100kHz to 110 MHz sweeper AM/FM 1970 vintage) and my old Tek 2235 scope (100MHz dual channel).
Using a 50 Ohm coax, and two T (to put 50 Ohm termination at each end) I plugged the scope on the gen.
I set the scope to have a nice view of the signal, and used the attenuator to have a big output without distortion of the sine wave (if you go beyond the "maximum allowed" by HP the sine wave get distorted on the positive half.
Then my wife came to see me and she played with the frequency adjust wheel. To my surprise, the scope reduced the amplitude displayed. The more you go up, the more the signal is faint on the screen. To the point that the scope loose sync. Of course the HP output indicator does not blink and stays proudly at the value I've set.
So I re-do the experiment starting at the minimum of the upper band (10 to 110 MHz) and the behavior was the same.
As my scope is 100 MHz bandwidth I thought that the display was around -3dB when the signal went to or near 100 MHz but the figure is much worse. (around 10 to 20 dB)
So how can I tell if it's the generator which fails or if the scope is too old ...
What do you think ? Could it be the coaxial wire attenuating the signal ?

ppppenguin · 10th Dec 2009, 4:03 pm

You must not use a termination at the generator end of the cable. The source impedance of the generator is 50R. Using an extra termination at the generator will cause a mismatch. This is very likely to give frequency response errors.

Some 'scopes (such as the 2465) have a switch to terminate the input in 50R. This is better than using a T and external termination but it's not very important at 100MHz.

peter_scott · 10th Dec 2009, 4:05 pm

Hi Georges,

I think the 8601 has 50 ohm output impedance so you shouldn't be adding another 50 load and T piece at the output.

Could you try the measurement again with your cable terminated with T and 50 ohm at the scope input but directly connected to the 8601?

Peter

Oops! Geoffrey beat me to it.

newlite4 · 10th Dec 2009, 4:05 pm

It could well be the capacitance of the coupling George. Equipment of this quality and stability is rarely to blame. Try to keep the coupling short and ensure that impedances match correctly.
Neil

Oldtestgear · 10th Dec 2009, 4:57 pm

The correct method to connect the 8601A to the 2235 with a T piece and terminator at the i/p to the 2235. This should then work correctly as the 8601A is seeing a 50R input rather than the 1Meg of the scope.

Is your terminator a good quality one or a cheapie from am old coax computer network? This could have an effect at high frequencies it was not intended to see.

Running without a terminator will not cause any stability problems for most signal generators but the output will be uncalibrated and higher than specified. This would allow you to check the 8601A output when connected directly to the 2235. If the output on the scope is still down then you need to borrow a scope to check if it is the 8601A or 2235 that is the problem.

The 2235 is a good reliable old scope but with two uncalibrated bits of test equipment it is always useful to get an independent check. It isn't impossible that both have problems. Been there & got the scars from that scenario.

Phil

georgesgiralt · 10th Dec 2009, 9:13 pm

Ahem...
I asked a question and I shouldn't...
First, the cable is, maybe, not 50 Ohm. Nothing written on it. I had some cables of various length, and can't find them ! So this one is questionable.
Next the terminations, they where in the scope pouch, are all different and can't tell who's the maker (nothing written on them). Maybe el cheapo computer stuff.
Last, the operator .... The worst part of the problem. When you've a scope with a button labelled "bandwidth limit 20 MHz" you've to check it is not depressed if you want to display 100 MHz signals... and re-read the manual to make sure you're correct...
Still if I cross the frequency range from bottom to top, the trace on the scope has definite minimals and maximals. They're repeatable and occur at very precise frequencies. I suspect that the setup act as a filter and what I see is actually the attenuation of the resonant frequency and it's harmonics. But the level does not drop as dramatically as it was ...
The only complaint left to the generator is that it looks like the display is not correct. It seems I've to be *above* the the goal frequency to get it right. I've to check that and use the sweep and the scope to see the IF bandwidth of one of my radio.
I'll try to borrow a counter next week and check this. For now I've to extinguish a well known broadcast to check the calibration ... This time, I'll triple check the dials and setting in order to reduce the amount of dumb question to this forum ;-) One never change .....
Thank you for your help and pointers to the most obvious mistakes into what I had written ;-)

Brian R Pateman · 10th Dec 2009, 10:27 pm

Don't worry about asking dumb questions Georges, we all do it and it's why this forum exists!

You are not the first person to fall fouls of the bandwidth switch. My 475 has one too.

With the lengths involved in test equipment hook ups the cable impedance is not too critical, the termination is the all important thing. Another point is to check that the plugs are fitted properly and the screen is adequately terminated. This sort of problem can give attenuation effects at certain frequencies.

Regards,

Station X · 11th Dec 2009, 10:10 am

I have the 8601A service manual as a pdf should you need it.

georgesgiralt · 11th Dec 2009, 4:17 pm

Hello Graham,
Thanks but I've got one.
I've got a nice book with it. Around 200 pages full of pictures and parts table. There is one also on Bama.
seems mine was out of the factory around 1970. so it is one of the last.
Next week I should have a counter from job. Very precise.
I hope I'll be more clever with the counter....

Station X · 11th Dec 2009, 5:14 pm

I think my manual came from the HP archives.

I've yet to use my 8601A in anger, but I tested it with a scope and counter and don't remember any problems with varying amplitudes.

jimmc101 · 11th Dec 2009, 6:27 pm

Could I suggest you try changing the length of the cable.
If this changes the frequency of the maxima and minima then the termination at the 'scope end is the problem.
(The frequencies should scale by roughly the inverse ratio of the length. eg 1.5 x length gives 2/3 x frequency)

As an aside, the input impedance of a '465 is quoted as 1M ohm // 20pF (Sorry I can't find the 2235 circuit).
A lot of people don't realise that this idoes not apply at tens of MHz.

Looking at the input circuit (attached) there is a 51 ohm damping resistor in series with the input connector.
Ignoring the capacity of the connector itself this means that the input impedance is actually 51 ohm series (1 M ohm // 20pF).
Negligible effect at low frequencies, but at 100 MHz this is equivalent to an input impedance of approx 180 ohm // 15pF

Jim

georgesgiralt · 11th Dec 2009, 10:47 pm

Hello Guys !
I'm quite sure the generator is fine and perfect (my only doubt is a slight discrepancy between the displayed frequency and the displayed one, I'll know for sure on monday night....).
as for the cables, the one I use is very long, as I can't find the cable I got with the scope. I normally use the scope with it's Tektro probes. So the cables are buried somewhere...
Unfortunately I'm far from expert at the use of scopes and high frequencies. so I'd better check all buttons before complaining ;-)
Mods, I think this embarrassing thread (for me) can be closed now ;-)

10th Dec 2009, 3:37 pm	#1
georgesgiralt Hexode Join Date: Jul 2008 Location: Toulouse, France Posts: 406	How could I tell? HP8601A RF Sweep Generator. Hi ! Yesterday, I was playing around with my "brand new" HP8601A RF sweep generator (100kHz to 110 MHz sweeper AM/FM 1970 vintage) and my old Tek 2235 scope (100MHz dual channel). Using a 50 Ohm coax, and two T (to put 50 Ohm termination at each end) I plugged the scope on the gen. I set the scope to have a nice view of the signal, and used the attenuator to have a big output without distortion of the sine wave (if you go beyond the "maximum allowed" by HP the sine wave get distorted on the positive half. Then my wife came to see me and she played with the frequency adjust wheel. To my surprise, the scope reduced the amplitude displayed. The more you go up, the more the signal is faint on the screen. To the point that the scope loose sync. Of course the HP output indicator does not blink and stays proudly at the value I've set. So I re-do the experiment starting at the minimum of the upper band (10 to 110 MHz) and the behavior was the same. As my scope is 100 MHz bandwidth I thought that the display was around -3dB when the signal went to or near 100 MHz but the figure is much worse. (around 10 to 20 dB) So how can I tell if it's the generator which fails or if the scope is too old ... What do you think ? Could it be the coaxial wire attenuating the signal ?

10th Dec 2009, 4:03 pm	#2
ppppenguin Retired Dormant Member Join Date: Dec 2003 Location: North London, UK. Posts: 6,168	Re: How could I tell ? You must not use a termination at the generator end of the cable. The source impedance of the generator is 50R. Using an extra termination at the generator will cause a mismatch. This is very likely to give frequency response errors. Some 'scopes (such as the 2465) have a switch to terminate the input in 50R. This is better than using a T and external termination but it's not very important at 100MHz.

10th Dec 2009, 4:05 pm	#3
peter_scott Dekatron Join Date: May 2003 Location: Edinburgh, UK. Posts: 3,274	Re: How could I tell ? Hi Georges, I think the 8601 has 50 ohm output impedance so you shouldn't be adding another 50 load and T piece at the output. Could you try the measurement again with your cable terminated with T and 50 ohm at the scope input but directly connected to the 8601? Peter Oops! Geoffrey beat me to it. __________________ http://www.nostalgiatech.co.uk https://www.youtube.com/c/PeterScott/videos Last edited by peter_scott; 10th Dec 2009 at 4:18 pm.

10th Dec 2009, 4:05 pm	#4
newlite4 Octode Join Date: Nov 2003 Location: Bath, Somerset, UK. Posts: 1,805	Re: How could I tell ? It could well be the capacitance of the coupling George. Equipment of this quality and stability is rarely to blame. Try to keep the coupling short and ensure that impedances match correctly. Neil __________________ preserving the recent past, for the distant future.

10th Dec 2009, 4:57 pm	#5
Oldtestgear Hexode Join Date: May 2008 Location: Devon, UK. Posts: 307	Re: How could I tell ? The correct method to connect the 8601A to the 2235 with a T piece and terminator at the i/p to the 2235. This should then work correctly as the 8601A is seeing a 50R input rather than the 1Meg of the scope. Is your terminator a good quality one or a cheapie from am old coax computer network? This could have an effect at high frequencies it was not intended to see. Running without a terminator will not cause any stability problems for most signal generators but the output will be uncalibrated and higher than specified. This would allow you to check the 8601A output when connected directly to the 2235. If the output on the scope is still down then you need to borrow a scope to check if it is the 8601A or 2235 that is the problem. The 2235 is a good reliable old scope but with two uncalibrated bits of test equipment it is always useful to get an independent check. It isn't impossible that both have problems. Been there & got the scars from that scenario. Phil

10th Dec 2009, 9:13 pm	#6
georgesgiralt Hexode Join Date: Jul 2008 Location: Toulouse, France Posts: 406	Re: How could I tell ? Ahem... I asked a question and I shouldn't... First, the cable is, maybe, not 50 Ohm. Nothing written on it. I had some cables of various length, and can't find them ! So this one is questionable. Next the terminations, they where in the scope pouch, are all different and can't tell who's the maker (nothing written on them). Maybe el cheapo computer stuff. Last, the operator .... The worst part of the problem. When you've a scope with a button labelled "bandwidth limit 20 MHz" you've to check it is not depressed if you want to display 100 MHz signals... and re-read the manual to make sure you're correct... Still if I cross the frequency range from bottom to top, the trace on the scope has definite minimals and maximals. They're repeatable and occur at very precise frequencies. I suspect that the setup act as a filter and what I see is actually the attenuation of the resonant frequency and it's harmonics. But the level does not drop as dramatically as it was ... The only complaint left to the generator is that it looks like the display is not correct. It seems I've to be above the the goal frequency to get it right. I've to check that and use the sweep and the scope to see the IF bandwidth of one of my radio. I'll try to borrow a counter next week and check this. For now I've to extinguish a well known broadcast to check the calibration ... This time, I'll triple check the dials and setting in order to reduce the amount of dumb question to this forum ;-) One never change ..... Thank you for your help and pointers to the most obvious mistakes into what I had written ;-)

10th Dec 2009, 10:27 pm	#7
Brian R Pateman Nonode Join Date: Jun 2005 Location: Western Lake District, Cumbria (CA20) - UK Posts: 2,136	Re: How could I tell ? Don't worry about asking dumb questions Georges, we all do it and it's why this forum exists! You are not the first person to fall fouls of the bandwidth switch. My 475 has one too. With the lengths involved in test equipment hook ups the cable impedance is not too critical, the termination is the all important thing. Another point is to check that the plugs are fitted properly and the screen is adequately terminated. This sort of problem can give attenuation effects at certain frequencies. Regards, __________________ Brian

11th Dec 2009, 10:10 am	#8
Station X Moderator Join Date: Jan 2003 Location: Ipswich, Suffolk, IP4, UK. Posts: 21,288	Re: How could I tell ? I have the 8601A service manual as a pdf should you need it. __________________ Graham. Forum Moderator Reach for your meter before you reach for your soldering iron.

11th Dec 2009, 4:17 pm	#9
georgesgiralt Hexode Join Date: Jul 2008 Location: Toulouse, France Posts: 406	Re: How could I tell ? Hello Graham, Thanks but I've got one. I've got a nice book with it. Around 200 pages full of pictures and parts table. There is one also on Bama. seems mine was out of the factory around 1970. so it is one of the last. Next week I should have a counter from job. Very precise. I hope I'll be more clever with the counter....

11th Dec 2009, 5:14 pm	#10
Station X Moderator Join Date: Jan 2003 Location: Ipswich, Suffolk, IP4, UK. Posts: 21,288	Re: How could I tell ? I think my manual came from the HP archives. I've yet to use my 8601A in anger, but I tested it with a scope and counter and don't remember any problems with varying amplitudes. __________________ Graham. Forum Moderator Reach for your meter before you reach for your soldering iron.

11th Dec 2009, 6:27 pm	#11
jimmc101 Heptode Join Date: Mar 2009 Location: Portsmouth, Hampshire, UK. Posts: 674	Re: How could I tell? HP8601A RF Sweep Generator. Could I suggest you try changing the length of the cable. If this changes the frequency of the maxima and minima then the termination at the 'scope end is the problem. (The frequencies should scale by roughly the inverse ratio of the length. eg 1.5 x length gives 2/3 x frequency) As an aside, the input impedance of a '465 is quoted as 1M ohm // 20pF (Sorry I can't find the 2235 circuit). A lot of people don't realise that this idoes not apply at tens of MHz. Looking at the input circuit (attached) there is a 51 ohm damping resistor in series with the input connector. Ignoring the capacity of the connector itself this means that the input impedance is actually 51 ohm series (1 M ohm // 20pF). Negligible effect at low frequencies, but at 100 MHz this is equivalent to an input impedance of approx 180 ohm // 15pF Jim Attached Thumbnails Last edited by jimmc101; 11th Dec 2009 at 6:33 pm. Reason: English

11th Dec 2009, 10:47 pm	#12
georgesgiralt Hexode Join Date: Jul 2008 Location: Toulouse, France Posts: 406	Re: How could I tell? HP8601A RF Sweep Generator. Hello Guys ! I'm quite sure the generator is fine and perfect (my only doubt is a slight discrepancy between the displayed frequency and the displayed one, I'll know for sure on monday night....). as for the cables, the one I use is very long, as I can't find the cable I got with the scope. I normally use the scope with it's Tektro probes. So the cables are buried somewhere... Unfortunately I'm far from expert at the use of scopes and high frequencies. so I'd better check all buttons before complaining ;-) Mods, I think this embarrassing thread (for me) can be closed now ;-)