11-08-2020, 02:55 PM
For better or worse I was one of the main designers of the T series mixers. Well the T16 anyway. I left Cox before the T8 but it used a fair bit of my work. Don't worry about telling distressing stories. In retrospect we made several major mistakes:
* The clicky buttons aren't very nice.
* The "incpot" spring loaded pots were not a good idea. We arrived at them after a lot of discussion on how to automate a mixer. The problem was reconciling the logical and mechanical positions of lots of controls. There are various solutions that really don't work. A continously rotating shaft encoder would have been ideal but these were expensive and too large at the time. Of several unsatisfactory options I think we chose the wrong one.
* We should have gone for bigger PCBs (we used 4U high) and 2 part edge connectors.
Before SCH phase was properly controlled any VT edit was a potential nightmare. I've done SCH measurements in various ways. You can digitise enough of the signal and do it in software. I've not done this though Tek probably did in their big all singing digital test sets. You can regenerate subcarrier and sample it. It's hard to get the sampling datum accurate enough but I've done it and it works. I thnk the Tek 1755A etc does it this way. For development work I needed to make absolute measuements, not depending on SC regen etc. I've built special modes into my SPGs which make a non-standard waveform where I can overlay the SC and H datum points on a scope. I claim, with a certain amount of hand waving, to get to within a couple of degrees with confidence. I have also computed SCH accurate waveforms to within 1 degree. The DAC reconstruction filter is often less accurate than that!
This is all very legacy stuff now. Back c1990(?) I wrote an article about SCH measurement for one of the broadcast magazines. This was basically advertorial for Hamlet - I was doing some work for them at the time. Less said about all that the better.
The Cox 350 clock had a special "ABC" keyboard option as well as the common QWERTY. It was compact which was useful on a desk but the main reason was union demarcation rules at Thames. To use a QWERTY keboard you had to be a typist. The ABC version could be used by the production assistant. From what I remember of the 350 design, getting the circle to look reasonable relied far too heavily on the tolerances of a lot of transistors. When I looked at it, my comments about the Thames designers were not printable.
Would love to hear more of your stories. Also about you and where you worked. If you don't want to go public on GVR please contact me by email or private message.
* The clicky buttons aren't very nice.
* The "incpot" spring loaded pots were not a good idea. We arrived at them after a lot of discussion on how to automate a mixer. The problem was reconciling the logical and mechanical positions of lots of controls. There are various solutions that really don't work. A continously rotating shaft encoder would have been ideal but these were expensive and too large at the time. Of several unsatisfactory options I think we chose the wrong one.
* We should have gone for bigger PCBs (we used 4U high) and 2 part edge connectors.
Before SCH phase was properly controlled any VT edit was a potential nightmare. I've done SCH measurements in various ways. You can digitise enough of the signal and do it in software. I've not done this though Tek probably did in their big all singing digital test sets. You can regenerate subcarrier and sample it. It's hard to get the sampling datum accurate enough but I've done it and it works. I thnk the Tek 1755A etc does it this way. For development work I needed to make absolute measuements, not depending on SC regen etc. I've built special modes into my SPGs which make a non-standard waveform where I can overlay the SC and H datum points on a scope. I claim, with a certain amount of hand waving, to get to within a couple of degrees with confidence. I have also computed SCH accurate waveforms to within 1 degree. The DAC reconstruction filter is often less accurate than that!
This is all very legacy stuff now. Back c1990(?) I wrote an article about SCH measurement for one of the broadcast magazines. This was basically advertorial for Hamlet - I was doing some work for them at the time. Less said about all that the better.
The Cox 350 clock had a special "ABC" keyboard option as well as the common QWERTY. It was compact which was useful on a desk but the main reason was union demarcation rules at Thames. To use a QWERTY keboard you had to be a typist. The ABC version could be used by the production assistant. From what I remember of the 350 design, getting the circle to look reasonable relied far too heavily on the tolerances of a lot of transistors. When I looked at it, my comments about the Thames designers were not printable.
Would love to hear more of your stories. Also about you and where you worked. If you don't want to go public on GVR please contact me by email or private message.
www.borinsky.co.uk Jeffrey Borinsky www.becg.tv
