Farnell signal generators

[WaveyDipole]

Is anyone familiar with these? For example the PSG1000 or the ESG1000?

The ESG1000 seems to be very similar to the PSG1000 but without the SINAD meter and a few other minor differences:

- sweep duration is fixed and cannot be varied
- no sellcall mode
- no DC power option
- no facility to alter the synthesizer rate - internal 1kHz tone only
- no mVemf or µVemf buttons
- no RAD PM button

There may be other as yet undiscovered differences, but otherwise the unit seems very similar in layout and function to the PSG1000.

I would specifically like to further understand the function of the DCFM button and the kHzFM button. From the manual it appears that the generator is capable of producing both AM and FM signals? I am uncertain as to the function of the DCFM button, but the manual states:

Originally Posted by Farnell Instructions/Service Manual for PSG1000

a) For modulatin level the three possible terminating keys are % AM, kHz FM and RAD PM. The resulting level is displayed in the four digit modulation level window.

For % AM the displayed resolution is 0.1%, and 0.01 radian for RAD PM.

With kHz FM, the displayed resolution varies depending upon the carrier frequency and the modulation level (see specification).

"b) The modulation circuits may be d.c. coupled with %AM, kHz FM by selecting DCFM.

In the % AM mode the isolating capacitor in the modulation input path is shorted out so that all the modulation paths are DC coupled.

When selected with kHz FM or RAD PM the synthesizer loop integrater is held at the last tuning voltage thus disabling the correction loop. The unlocked carrier can now be frequency modulated with a d.c. voltage at the modulation input.

By selecting MOF OFF the modulation signal is isolated from the carrier circuits thus removing the modulation. Additionally, MOD OFF deselects DC FM (if selected), permitting re-locking of the carrier.

The reference to Radians can be disregarded as this unit does not have that function, but I had supposed that pressing kHz FM would put the generator into FM modulation mode. Instead, it appears to be stuck in AM mode and the kHz FM button does nothing except for showing 'ignore' on the display. I am unfortunately none the wiser as to the purpose of the DC FM button.

The other puzzle is the sweep button. This does perform a very slow sweep. On the the PSG1000 apparently this can be varied between 2 and 200 seconds, but on this unit, it appears to be fixed as there is no opportunity to input the number of seconds despite the button legend. It also puzzles me that there is no trigger output which means that the sweep cannot synchronised and viewed on a scope?

Finally, despit the instrument telling me that its GPIB address is 16, there is no response from the GPIB interface so it seems there are multiple problems which require investigating. However, as a staring point, it would be useful to at least understand how the kHz FM and DC FM buttons *should* work.

Any information, or even a copy of the ESG1000 manual would be appreciated.

[IanG3XYV]

I have a PSG1000 but haven't used it in any of the modes you are asking about I'm afraid.

I don't use it that often but recall that it is not the easiest of sig gens to set up. I quite often get that "ignored" message when using it after a while. You need to be clear if you are in "carrier" or "modulation" mode when entering keyboard data although some keys work independently. Not tried GPIB.

I have the Instructions/Service Manual for the PSG1000 which you are welcome to if you let me have your email. It's too large to attach here. It appears to be very similar to the extracts you included in your post so it may not add to your picture but it does have circuit diagrams and parts lists.

It's a good quality, compact synthesised RF sig gen and I wish you success with yours which I imagine is very similar to mine.

Ian

[WaveyDipole]

Ian,

Yes, you describe the push button operation accurately. The Int and Ext and Mod On keys work independently. For most of the others it seems one has to be in the right mode and with the cursor in the correct display, although the manual indicates that %AM and kHz FM should operate in either Display Carrier or Modulation mode. But yes, one has to get the combination right otherwise one is greeted with 'ignore'. I still don't know what the button with the triangle is at the top right.

I will PM you my e-mail address for a copy of that manual. It may well be the same one as I have but no harm having a look in case it adds something. I couldn't find one specifically for the ESG1000 anywhere.

I checked the power rails today and all seem correct. I was only able to check the 5V rail driven by U2. Not sure where the one driven by U1 connects yet. Will probably have to disassemble the rear to get to it. I also tried re-seating some of the chips, but I might have caused further problems by doing that AM modulation is now lost as well. Only the carrier is present but both %AM and kHz FM now say 'ignore'.

I am puzzled by GPIB problem. I have confirmed that there is power to all the related ICs and I substituted both SN75160 and SN75161 (just so happens I have known good spares) but no joy. The next IC back is the UPD7210C GPIB controller chip for which I do not have a spare. The 8085 processor must be working since we have a display and I have confirmed power, 2MHz clock and ALE signals present.

I did a bit of probing around the modulation control (CT3) and was able to confirm the operation of the frequency divider ICs and trace the signal to the output coupling capacitor but it then doesn't seem to follow the diagram to pin 10 of U301. Unfortunately I ran out of time to probe further.

I really need to get my head around how this all works together. Clearly the buttons work as there is a click and some light up as functions are enabled and others respond by displaying 'ignore'. Either way there is a response which suggests to me that the processor is detecting the press and is responding. Quite how the processor controls the functions and whether they are responding correctly I am not sure of yet and need to study this further.

While not a HP or a Tek, the unit is built well enough, certainly much better than many plastic cased instruments of today. It might just be worth fixing if I can get my head around it!

[greenkit]

I also have a PSG1000 that behaves oddly: in my case I can't select AM and get the 'ignore' message when trying to select the % mod. It's been like this since I've had it. FM is fine, as are any of the other generator functions I've needed. I've never used the GPIB. One niggle is that the output protection function is very sensitive and activates a bit too easily, turning the RF off.

As far as I know the DC coupled modulation facility was included to cater for low-speed data inputs.

[IanG3XYV]

You are certainly getting into the works which I haven't needed to do. I have used it in AM and FM mode for alignment and testing of a number of valve and solid state radios but haven't used it to sweep.

You are right it is well screened and has a good attenuator capable of providing sub microvolt signals for checking sensitivity of communications receivers.

I'll try the GPIB next time I have it on the bench for signs of life. I have a NI USB adapter which I can use to send basic commands to devices. I'm assuming you have something similar.

I'm emailing the manual I have but suspect it may be the same as yours.

Ian

[WaveyDipole]

Originally Posted by greenkit

I also have a PSG1000 that behaves oddly: in my case I can't select AM and get the 'ignore' message when trying to select the % mod. It's been like this since I've had it. FM is fine, as are any of the other generator functions I've needed. I've never used the GPIB. One niggle is that the output protection function is very sensitive and activates a bit too easily, turning the RF off.

As far as I know the DC coupled modulation facility was included to cater for low-speed data inputs.

Thanks for the comment. It sounds like your unit has a similar problem but the opposite effect. I will keep an eye out regarding the sensitivity. If the RF output switches off for any reason I will know what it is. If it trips, will the unit need to be power cycled, or does it resume one the overload (perceived or otherwise) is removed?

Originally Posted by IanG3XYV

You are certainly getting into the works which I haven't needed to do. I have used it in AM and FM mode for alignment and testing of a number of valve and solid state radios but haven't used it to sweep.

That was my intended usage and it would be useful to have both AM and FM.

Originally Posted by IanG3XYV

I'll try the GPIB next time I have it on the bench for signs of life. I have a NI USB adapter which I can use to send basic commands to devices. I'm assuming you have something similar.

Thank you. I would be interested to hear how you get on. I had hoped that the FM functions that I can't get to work via the front panel buttons might at least work via GPIB, but with GPIB being dead, I am unable to test that.

Originally Posted by IanG3XYV

I'm emailing the manual I have but suspect it may be the same as yours.

I saw your e-mail and yes it was, but thank you for taking the time and trouble. Much appreciated.

if I do figure anything out I will post it back on here.

[WaveyDipole]

While probing around today I noticed something a little odd on the 8055 CPU data bus. I was probing at the CPU side data pins pins on the GPIB talker chip (pins 12 - 19). While some appeared to show clean square digital pulses, a couple of the pins showed some artefacts. Can anyone shed some light on why this might be?

[IanG3XYV]

It's a while since I poked around on a CPU bus but I don't think you have anything to worry about. The first shows a bit of noise which you will always get through coupling and loading but it is clean around the logic threshold. The second I suspect is a line going tri-state or high impedance from a "1" level. The capacitance of the pcb trace keeps it high momentarily but it decays exponentially to a "0" as it discharges.

If you are able to set frequencies and RF levels the CPU would appear to be functioning properly. Perhaps there is a problem with the switch matrix? Not sure if this explain the GPIB lack of operation.

[WaveyDipole]

Fair enough. It is possible to set the carrier frequency and level and the display is working. There is a response to all key presses so I agree, the CPU must be running.

Through trial and error I have discovered that GPIB does actually work! For some reason the #55 sequence which is supposed to display the GPIB address always shows 16 instead of the current GPIB address. The actual address was set to something else. I have confirmed that the instrument can communicate on the address that is set with NN#55 sequence where NN is the desired address. The address is stored between power cycles but if not set to 16, one has to remember what it is set it to! I have now set it to 16.

Since the manual states that the terminator is a carriage return, it is necessary to set the terminator as carriage return only on the GPIB adapter. The instrument then accepts commands as described in the manual. Also as described, commands are executed just as though the corresponding key had been pressed.

Unfortunately %AM and kHzFM still do not work. They behave exactly the same as pressing the corresponding button and just return 'ignored'. This would seem to rule out problems with the keyboard membrane and the decoding of key-presses by the processor and would tend to suggest that the problem is with how the processor executes the command, or at least with how that command translated from the processor to the RF generator unit. There are a number of 74xx series and the odd CMOS ICs that form the switching logic between the CPU and the RF generator. That is where my next efforts will be focused.

Still a way to go, but a little progress made.

[greenkit]

For info regarding the RF trip: the 'RF Off' indicator comes on, just hit the membrane switch to reset.

[WaveyDipole]

greenkit, noted. Thanks.

[WaveyDipole]

It seems that the only problem here was getting the keying sequence correct! It is perhaps natural to think that pressing the function key (%AM or kHzFM) will select the corresponding modulation type after which one might proceed to adjust the modulation parameters as desired. However, after a bit of head scratching and repeated re-reading of the manual it eventually became apparent that this Farnell generator expects the modulation level to be entered first, followed selecting the required modulation function.

For anyone else coming across this issue, these are the necessary steps to take:

a) select DISP MOD mode by pressing the DISP MOD key
b) key in a value to represent the modulation level and then press %AM or kHzFM

The %AM or kHzFM mode button should then light up and the modulation level should be displayed on the 4-digit right hand side display. Pressing the %AM or kHzFM button on its own displays 'IGnorEd' because it hasn't been preceded with a value to allow the generator to set the desired modulation level.

For example, to set AM modulation at 50%, make sure the generator is in DISP MOD and key '50' on the numeric keypad followed by the %AM key. The %AM key should light up and 50.0 should appear on the right hand display. The output should be a carrier 50% AM modulated with the internal or external modulation source. Likewise, keying '12.5' followed by kHzFM key should light up the kHzFM key, 12.5 should appear on the right hand display, and the internal or external modulation source should be applied to the carrier with a peak deviation of 12.5kHz.

The information is actually present in the manual in section 5.4 (a), but might not seem very clear when the section is read in isolation. However, when one bears in mind the information already contained in the preceding sections 5.1 (b) and 5.2 (a+b) it then becomes apparent that the Modulation Level control follows the same keying pattern, i.e value followed by function key.

The permitted values to use for each modulation type are contained in the specification section of the manual.

I hope this information will help others encountering the same difficulty.

[IanG3XYV]

Thank you WD for taking the trouble to set that down in writing. I think I came to that conclusion in my occasional uses of my PSG1000 but documentation was never my forte. I just remember noting to myself that this user interface is not intuitive. Well done!

[Radio Wrangler]

DC coupled modulation has a couple of uses that aren't terribly obvious:

DC FM can be used to phase-lock a sig gen onto a signal which can allow the phase noise to be mixed down to the range of an audio spectrum analyser (or soundcard) assuming the sig gen is cleaner than what's being investigated. This can be done with just a packaged mixer, say from mini-circuits. The FM deviation control is used to set the loop gain and hence the loop bandwidth.

DC AM can be used as an alternative to having an external levelling detector but may need circuitry in addition to just the detector unit.

David

[WaveyDipole]

David, thank you for explaining that. These are advanced applications that I may never actually make use of although one never knows when such features might come in handy.

I would still like to clarify the purpose of the sweep. Using the HP3314A, I can inject the signal output into an IF can, connect the Trig output to channel 2 of the oscilloscope and use it as a trigger signal, then set a start and stop frequency (e.g. 400-500MHz) and sweep the IF can. With the oscilloscope time-base set to show exactly one trigger pulse across the width of the display I can view a trace that shows me the peak with each vertical graticule line showing me exactly 10kHz increments. The result is shown in the photo.

This Farnell ESG1000 has a sweep time of 2 seconds to 200 seconds compared to the HP3314 default time of 10ms. There is no Trig output so there seems to be no possibility of synchronising the oscilloscope to the generator and replicating the result obtained from the HP3314A. While peaking can, of course, be done in other ways, e.g. by observing the amplitude of the waveform at the required frequency or the output voltage on a voltmeter, or simply by listening to the volume of the demodulated tone, I am curious what such long and un-synchronised sweep times might be useful for?

Interestingly I couldn't replicate the result in the photo when using the cheap Feeltech FY6900 either despite the fact that it does have a 'Sync Out' socket and a Trig socket and sweep time adjustable down to 10ms. The Trig socket is quiet and the Sync Out socket does not generate a steady pulse like the HP3314, but a square wave that appears to sweep in rhythm with the sweep signal. In fact it gave me some very weird and rather square-ish artefacts suggesting that the sweep might be accomplished in rather course steps.

[Radio Wrangler]

You can use slow sweeps without needing to trigger if you use a spectrum analyser set to 'max hold' instead of a scope. It's kludgy but if you haven't a tracking generator then it gets you by.

There are all sorts of measurement tricks from out and out bodges to some really rather subtle ones.

You build up a portfolio as you go along.

No matter how well equipped the lab you work in, you never have everything so sometimes you have to get inventive, and then there are the cases where you're trying to measure something no-one's really had to do before. I've been quite proud of some of my spur of the moment bits of creativity (and made sure no-one spotted just how bodgy some others were...)

David

[greenkit]

Wavey, thanks for that information, very useful and it has solved my problem too! I will post myself an appropriate note on top of the genny. My regular driver is an old TF2008 that's more intuitive, but the PSG1000's stability and resolution do make it very useful.

[WaveyDipole]

Originally Posted by Radio Wrangler

You can use slow sweeps without needing to trigger if you use a spectrum analyser set to 'max hold' instead of a scope. It's kludgy but if you haven't a tracking generator then it gets you by.

David, I don't have a decent spectrum analyser, although I have just received a TinySA to play with, but thanks for the information.

Originally Posted by greenkit

Wavey, thanks for that information, very useful and it has solved my problem too! I will post myself an appropriate note on top of the genny. My regular driver is an old TF2008 that's more intuitive, but the PSG1000's stability and resolution do make it very useful.

greenkit, I am glad the information helped. I had been on the lookout for a 2022 but came across this ESG1000 in the meantime. It is handy to be able to cover all bands from LW through to UHF in one unit and to be able to set the carrier amplitude in mV as well as in dBm and go down into μV and dBμV. It is also sufficiently compact to just about fit on my shelf, so this has worked out well enough in the end. If I come across a 2022D at the right price in the future, then I might decide to "upgrade", but the ESG1000 will certainly be useful for the present.