|
Vintage Computers Any vintage computer systems, calculators, video games etc., but with an emphasis on 1980s and earlier equipment. |
|
Thread Tools |
9th Nov 2004, 9:23 pm | #1 |
Retired Dormant Member
Join Date: Mar 2003
Location: Bishop's Waltham, Hants, UK.
Posts: 939
|
Thyratron pulse generator (for Sam)
Sam,
I have a simple thyratron pulse generator (see question in this thread https://www.vintage-radio.net/forum/...read.php?t=154) running on the bench, using an EN91, and four other components, to produce a negative going pulse (well, spike really ), of about -10V amplitude. The downside is it needs a -ve bias supply, and the cheat is I used my existing stabilised HT supply . However, I do understand the maths of the timing, and it does produce reasonably repeatable results. I will attempt to write it up and post it on my website in the next few days, and if I can get my Tek 535A to work again (it expired in the middle of the experiment ), I'll try and get some waveform pictures. The test model is currently running at about 25Hz, as that is the slowest the pathetic modern Philips scope will trigger at (the Tek has no problem at low speed, but I will need my proper scope camera to take the picture, then I'll have to scan the Polaroid, there are some things that digi cameras just won't do!) but I have had it running as low as 1Hz. Jim. Last edited by Paul Stenning; 29th Dec 2004 at 3:23 pm. |
9th Nov 2004, 9:44 pm | #2 |
Hexode
Join Date: Apr 2003
Location: Higham-on-the-Hill, Leicestershire, UK.
Posts: 338
|
Re: Thyratron pulse generator (for Sam)
Wow! Thanks Jim! This idea of mine certainly seems to be geting a lot of people thinking and searching! Maybe I should start to finalise details, buy/scrounge components, and get properly started soon!
But the pulse/spike sounds perfect. Being negative may mean a few more alterations to Wynn-Williams' circuit, but I should get a working ring up and running before worrying about connecting them together! I think a non-lethal power supply will have to be my first priority though! Thanks to everyone who has made suggestions/given information. I will have to work out a way to keep everyone posted. I think I have some webspace on the Uni network somewhere... Sam
__________________
Can he lead a Normal life, Doctor? No, he will be ... an Engineer. |
9th Nov 2004, 10:04 pm | #3 |
Retired Dormant Member
Join Date: Mar 2003
Location: Bishop's Waltham, Hants, UK.
Posts: 939
|
Re: Thyratron pulse generator (for Sam)
Stage two (for tomorrow lunchtime) was to add a triode section, so you can have either a positive or negative going pulse - should square the pulse up a bit as well. The type of triode depends on what I have in the junk box, and what I can get away with bias wise.
Jim. |
9th Nov 2004, 10:51 pm | #4 |
Hexode
Join Date: Apr 2003
Location: Higham-on-the-Hill, Leicestershire, UK.
Posts: 338
|
Re: Thyratron pulse generator (for Sam)
Thanks once again Jim. Electronics isn't my forte (or I suppose I would have gone into the electronic stream )! I am learning as I go here! A bit different from op-amps, and K-map logic minimisation! Maybe the power electronics we do later in the year will help.
It is immeasureably more fun and interesting though! Sam
__________________
Can he lead a Normal life, Doctor? No, he will be ... an Engineer. Last edited by Station X; 25th Jan 2005 at 10:23 am. |
10th Nov 2004, 12:04 am | #5 |
Retired Dormant Member
Join Date: Mar 2003
Location: Bishop's Waltham, Hants, UK.
Posts: 939
|
Re: Thyratron pulse generator (for Sam)
Sam,
I can't find a way to get the diagram from my schematic programme (RIMA) into Frontpage to publish it, so it looks like I'll have to scan the picture. Valve for the ouput stage is looking like a 6SL7 or 6SN7, as I have a qauntity of each, and they can be cut off with somewhere in the region of -15V on the grid (Va=250V), so I won't have to seriously re-design the bias supply ( I'm currently using -10V, which can be easily derived from a 6.3V heater winding). I'm considering reducing the oscillator supply voltage to 150V, as this will fit in with a simple gas discharge type regulator (for example, a VR150/30, which is readily available from the junk box), there are two advantages to this: 1. Oscillator has a fairly stable supply - the frequency of the oscillator varies with input voltage (obvious from the maths) 2. Lower oscillator voltage will automatically decrease the frequency, allowing the use of smaller values of C and R. The downside is that the oscillator grid bias becomes more critical. I think I need to work on plotting the characteristics of the EN91 a little more, as from the graph I have (Brimar valve manual No8 ), is not very accurate for low bias values. Jim. Last edited by Station X; 25th Jan 2005 at 10:27 am. Reason: Import |
10th Nov 2004, 1:07 am | #6 |
Retired Dormant Member
Join Date: Mar 2003
Location: Bishop's Waltham, Hants, UK.
Posts: 939
|
Re: Thyratron pulse generator (for Sam)
|
10th Nov 2004, 8:31 am | #7 | |
Retired Dormant Member
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
|
Re: Thyratron pulse generator (for Sam)
Quote:
|
|
10th Nov 2004, 8:46 pm | #8 |
Retired Dormant Member
Join Date: Mar 2003
Location: Bishop's Waltham, Hants, UK.
Posts: 939
|
Re: Thyratron pulse generator (for Sam)
Jeffrey,
that's what I did in the end, but I would have preffered to insert a bitmap into the webpage, rather than a PDF which people have to open seperatly, still I shoudn't complain, the software was free, and came with valve libraries in it (I had to modify one to make a thyratron though ). Sam, I've done a little more dvelopment today, and have settled on a circuit for the oscillator stage, but I didn't do any work on the output part. I'll write (type?) it up later. Jim. |
11th Nov 2004, 12:18 am | #9 |
Retired Dormant Member
Join Date: Mar 2003
Location: Bishop's Waltham, Hants, UK.
Posts: 939
|
Re: Thyratron pulse generator (for Sam)
|
12th Nov 2004, 9:44 pm | #10 |
Retired Dormant Member
Join Date: Mar 2003
Location: Bishop's Waltham, Hants, UK.
Posts: 939
|
Re: Thyratron pulse generator (for Sam)
Sam,
I've done alittle more development, and now have a working 1Hz generator, with a narrow (15mS) 200V pulse output! It uses three valves, and not a lot else, the diagram will follow soon. The down side is it is relativly unstable (15ms jitter, giving a cumulative error of 22 minutes per day at worst - not much worse than the qouted stability of the mains). I don't think it is practical to take the free running design much further. The current circuit does however, synchronise to an incoming pulse very easily. I have done some reading on thyratron frequency dividers, and think it may be more practical to synchronise to the mains. I'll do a little work on this idea next week (the toys are at work at the moment ). I have also found a diagram for a thyristor (SCR) ring counter, which I think I can adapt for either some of the EN91's, or some G240/10D cold cathode trigger tubes I have in stock - another experiment for next week. Jim |
12th Nov 2004, 10:01 pm | #11 |
Hexode
Join Date: Apr 2003
Location: Higham-on-the-Hill, Leicestershire, UK.
Posts: 338
|
Re: Thyratron pulse generator (for Sam)
Thanks for keeping me posted Jim. I am still thinking, rather than doing!
Hopefully the ofd Farnell PSU I bought from eBay will turn up soon so I can start experimenting! Reading Winn-Williams' paper again, he talks about the critical (grid) voltge. I assums this is where the thyratron starts to conduct, and where the control curve flattens out at a low anode voltage. That was another thought I had of obtaining the pulses - somehow fron the 50Hz of the mains. Either by using a scale-of-50 circuit, or a decatron (would that work ) I just keep being sidetracked with problem sheets and real-time computing. Grr Sam
__________________
Can he lead a Normal life, Doctor? No, he will be ... an Engineer. |
12th Nov 2004, 10:56 pm | #12 | |
Retired Dormant Member
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
|
Re: Thyratron pulse generator (for Sam)
Quote:
|
|
12th Nov 2004, 11:13 pm | #13 | |
Octode
Join Date: Oct 2003
Location: Newbury, Berkshire, UK.
Posts: 1,770
|
Re: Thyratron pulse generator (for Sam)
Quote:
__________________
Chris |
|
12th Nov 2004, 11:21 pm | #14 |
Moderator
Join Date: Jun 2003
Location: Oxford, UK
Posts: 27,740
|
Re: Thyratron pulse generator (for Sam)
I believe most modern clock radios are locked to the mains frequency, and are very accurate. Most have a quartz backup timer driven by a PP3, and these are noticeably less accurate in my experience - in fact, I never bother with the backup battery
Best regards, Paul |
13th Nov 2004, 12:19 am | #15 |
Retired Dormant Member
Join Date: Mar 2003
Location: Bishop's Waltham, Hants, UK.
Posts: 939
|
Re: Thyratron pulse generator (for Sam)
New circuit has been uploaded:
www.g1jbg.co.uk/thyrpul.htm As regards mains frequency, the allowable error works out to 14.4 minutes per day, but the average is usually kept to 50Hz. The maximum permissable error is +/- 1% (form National Grid website). I have a couple of synchronous clocks here, and they tend to show an error of +/- 2minutes throughout the day, but are normally accurate at the same time each day (error depends on the prevailing load on the network, but averages to zero over 24 hours). In the early days, frequency control was very poor, until the Warren clock company wanted to sell a cheap synchronous clock in the States. To make it attractive, clock needed to be accurate, so they manufactured an accurate impulse master clock, which was fitted in the power station, and used an ingenious mechanism to compare the frequency of the power with the master clock. A synchronous motor drives a pointer around a dial, and coupled to this via a slipping clutch (friction coupling) is a second pointer. Attached to the second pointer is a star shaped wheel, into which a solenoid engages with each pulse from the mster clock. If the mains frequency is accurate, both pointers will have moved round one space on the star wheel when the solenoid operates, if, however, there is a difference, then when the solenoid operates, the slipping clutch will allow the two pointers to move apart, showing not only that there is an error, but also its magnitude and direction. The speed of the generators can then be altered to bring the pointers back into alignment. We still have a more modern one of these units in our generator house at work, which works on the same principle, but takes a 1 second impulse from the master clock. When it was installed (late 1960's), the master clock was a pendulum type impulse clock (GPO clock No36?), but the system now runs from a radio clock synchronised to MSF Rugby. If you have a synchronous clock with a sweep (seconds) hand, it is interesting to note the variation over the day, when compared to a radio clock. Jim. |
13th Nov 2004, 10:13 am | #16 |
Hexode
Join Date: Apr 2003
Location: Higham-on-the-Hill, Leicestershire, UK.
Posts: 338
|
Re: Thyratron pulse generator (for Sam)
Remembering something our Electromechanics lecturer said last year, the mains frequence does vary very slightly as the load fluctuates on the grid (P and Q ballance ). I think it is one way for the grid to absorb these changes. However, the number of waves generated was fixed by government, so that a synchronous clock would't gain or lose over a day. Apparently one of the jobs at a power station/monitoring station was to keep an eye on this, and coming up to midnight speed up the generators to make sure the correct number of cycles had been generated!
Looking at the updated circuit, I reckon I can try and create my own versions. I have a few old american metal-cased triodes kicking around! Sam
__________________
Can he lead a Normal life, Doctor? No, he will be ... an Engineer. |
13th Nov 2004, 10:24 am | #17 |
Retired Dormant Member
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
|
Re: Thyratron pulse generator (for Sam)
I have a friend who works at National Grid Control. He designs the software that decides which generators run when. I'll ask him about their policy of frequency control.
|
13th Nov 2004, 10:32 am | #18 |
Retired Dormant Member
Join Date: Mar 2003
Location: Bishop's Waltham, Hants, UK.
Posts: 939
|
Re: Thyratron pulse generator (for Sam)
Sam,
after looking at the valve book last night, a 6J5 has the same characteristics as half of a 6SN7 - it's just that I have a plan for the other half A 6SL7 would work as well, but if you only use one section, the bottom of the pulse wopuld only drop to 9V instead of 1V. Both sections strapped together would give a lower limit of around 4.5V, Though 9V is not a real problem, you just lower the bias on the thyratrons a bit. Jim. |
13th Nov 2004, 10:34 am | #19 | |
Octode
Join Date: Oct 2003
Location: Newbury, Berkshire, UK.
Posts: 1,770
|
Re: Thyratron pulse generator (for Sam)
Quote:
__________________
Chris |
|
13th Nov 2004, 10:47 am | #20 |
Retired Dormant Member
Join Date: Mar 2003
Location: Bishop's Waltham, Hants, UK.
Posts: 939
|
Re: Thyratron pulse generator (for Sam)
All the machines will remain synchronised (three phase theory), but the frequency of the whole lot may vary. To bring machines on line, the generator is run up off load, and the voltage between the grid and the generator for each phase is measured (the old way was a light bulb btween the generator phase and the bus bar phase), when the voltage reduces to zero (machine in phase with grid), then the breaker is closed to connect the generator to the grid.
I would suspect there is a backup system as well - our distributed clock system at work has two Rugby receivers, plus one for DCF (Mainflingen, Germany), on top of this the clock receivers fre run to a high level of accuracy, and if all that fails, there is a main and standby crystal oscillator as well (5 levels of redundancy). Jim. Jim. |