Thanks for your responses gentlemen.
My initial hope was that someone simply might have 'been there before' and could simply say something like "change XYZ and ABC". However, it seems not. O.K., perhaps time for some cct. analysis; I was asked for drawings.
I've found a few moments at the end of a busy day: drawings now attached.
The first one is the overall & complete drawing of the calibrator cct. Since the whole cct. is too far small to readily analyze its operation, I've attempted to split out the relevant bits and add them as additional drawings but with these drawn to a larger scale.
First, the overall diagram: gives you the general idea:
A few explanatory comments about the cct. as drawn and how it's meant to work are required: an overview of the ccts. function follows:
Crystal controlled 1 MHz osc., amplifier and limiter are fairly straightforward. The limiter feeds a ÷10 stage to produce 100 kHz and that feeds another ÷10 stage to produce 10 kHz.
Drawings below refer:
An output at 1 MHz from the xtal. osc. is fed to the base of mixer transistor TR12 via a Schmitt trigger and a pulse shaper. Another input to the TR12 mixer base is from the wired-AND O/Ps from the two ÷10 stages. The sig. gen. actual selected freq. is fed to the emitter of TR12 via a wide-band amplifier. The O/P of the TR12 mixer is fed through a 2 kHz LPF (remotely situated on filter modules A7 and A8), since only the heterodyne audio beat note is required. This beat note is fed to the calibrator amplifier (on a separate module). The d.c. supply for the collector of TR12 is derived from that calibrator amplifier and is fed to TR12 collector via that LPF. The coils in that LPF have been changed as part of the total rebuild of the filter modules, A7 and A8.
Drawings below refer:
Note that a -15v supply is made available to the 1 MHz osc. when the cal. is switched on and to the 1 MHz position only. When the cal. switch is set to 100 kHz, a -15v. supply is only then connected to the 100 kHz divider and the original -15v. supply for the 1 MHz osc. Ditto the 10 kHz divider. So, three supplies from a common source but all separately filtered remotely (modules A7 and A8) from the calibrator cct. and separately switched, according to the setting of the front panel cal. switch.
Investigation of the fault with the 100 kHz divider (and the rest of the board) is hampered by the difficulty of gaining access to various component lead-outs with the board in the unit: the component density is very high with all the Rs. fitted vertically to the pcb. What also does not help in the practical investigation of the cct. operation is that the whole cct. is built around a mix of NPN and PNP transistors running from a -15v. supply rail.
I have tried a 'blanket repair' by replacing TR3, TR4, TR5, D3, D4 and C16 with 'known good' items of exactly the same type, but to no avail. The O/P from the limiter stage looks correct on the 'scope.
What I do get in terms of calibrator beat notes are beats at 1 MHz but only beats at 500 kHz intervals when the calibrator is set to the 100 kHz position. The 10 kHz divider stage is totally non-functional, but to avoid this causing the or an additional fault, its output to the TR12 mixer has been disabled by the temporary removal of R20: the feed resistor from this stage to the base of the mixer transistor TR12.
Overall, quite a complicated arrangement: I hope my overview analysis is adequate. Naturally I will attempt to answer any questions thus arising. And thank you for wading through that lot!
Al. / Aug. 4th., 2012 //
My initial hope was that someone simply might have 'been there before' and could simply say something like "change XYZ and ABC". However, it seems not. O.K., perhaps time for some cct. analysis; I was asked for drawings.
I've found a few moments at the end of a busy day: drawings now attached.
The first one is the overall & complete drawing of the calibrator cct. Since the whole cct. is too far small to readily analyze its operation, I've attempted to split out the relevant bits and add them as additional drawings but with these drawn to a larger scale.
First, the overall diagram: gives you the general idea:
A few explanatory comments about the cct. as drawn and how it's meant to work are required: an overview of the ccts. function follows:
Crystal controlled 1 MHz osc., amplifier and limiter are fairly straightforward. The limiter feeds a ÷10 stage to produce 100 kHz and that feeds another ÷10 stage to produce 10 kHz.
Drawings below refer:
An output at 1 MHz from the xtal. osc. is fed to the base of mixer transistor TR12 via a Schmitt trigger and a pulse shaper. Another input to the TR12 mixer base is from the wired-AND O/Ps from the two ÷10 stages. The sig. gen. actual selected freq. is fed to the emitter of TR12 via a wide-band amplifier. The O/P of the TR12 mixer is fed through a 2 kHz LPF (remotely situated on filter modules A7 and A8), since only the heterodyne audio beat note is required. This beat note is fed to the calibrator amplifier (on a separate module). The d.c. supply for the collector of TR12 is derived from that calibrator amplifier and is fed to TR12 collector via that LPF. The coils in that LPF have been changed as part of the total rebuild of the filter modules, A7 and A8.
Drawings below refer:
Note that a -15v supply is made available to the 1 MHz osc. when the cal. is switched on and to the 1 MHz position only. When the cal. switch is set to 100 kHz, a -15v. supply is only then connected to the 100 kHz divider and the original -15v. supply for the 1 MHz osc. Ditto the 10 kHz divider. So, three supplies from a common source but all separately filtered remotely (modules A7 and A8) from the calibrator cct. and separately switched, according to the setting of the front panel cal. switch.
Investigation of the fault with the 100 kHz divider (and the rest of the board) is hampered by the difficulty of gaining access to various component lead-outs with the board in the unit: the component density is very high with all the Rs. fitted vertically to the pcb. What also does not help in the practical investigation of the cct. operation is that the whole cct. is built around a mix of NPN and PNP transistors running from a -15v. supply rail.
I have tried a 'blanket repair' by replacing TR3, TR4, TR5, D3, D4 and C16 with 'known good' items of exactly the same type, but to no avail. The O/P from the limiter stage looks correct on the 'scope.
What I do get in terms of calibrator beat notes are beats at 1 MHz but only beats at 500 kHz intervals when the calibrator is set to the 100 kHz position. The 10 kHz divider stage is totally non-functional, but to avoid this causing the or an additional fault, its output to the TR12 mixer has been disabled by the temporary removal of R20: the feed resistor from this stage to the base of the mixer transistor TR12.
Overall, quite a complicated arrangement: I hope my overview analysis is adequate. Naturally I will attempt to answer any questions thus arising. And thank you for wading through that lot!
Al. / Aug. 4th., 2012 //
