EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: TheBay on May 08, 2022, 06:40:40 pm
-
I've finally got myself a Leo Bondar Precision GPSDO, the dual output version.
So far I'm really happy with the unit and I have one output configured for 10MHz and the other configured for 1MHz (For my Marconi 2955R)
However it's a little tiresome swapping over BNC's in the workshop to put the 10MHz reference on different items of test equipment.
And thought I'd get myself a 10MHz distribution unit, I had this in mind previously while I was looking at different GPSDO's and figured I'd just buy a Video Distribution Amplifier and change the impedance.
However, something I didn't consider or think about during my purchase of the GPSDO is whether it has a squarewave or sinusoidal output, looking at my Leo Bondar unit on the scope the output is squarewave. Which is not currently an issue for the test equipment I'm using it with, but it does make me wonder about using it with a Video distribution amplifier.
Every article or post I've seen where these have been used for this purpose have been used with a GPSDO that has a sinewave output, has anyone used one here on a GPSDO with a squarewave output or a Leo Bodnar unit? I'm just wondering whether the opamps used in these devices will be suitable.
It makes no real difference to me if the output from the video distribution is square or sinusoidal, nor is phase or isolation important at this stage. That can be a consideration later on.
For now I just want a 10MHz reference to my AWG, Frequency Counter, RF Gen etc. 4 outputs will do the lot but 6 would give me some spare.
I was looking at the Extron units like the MDA 4V, MDA 5V etc. I spotted a few Canford 90-541 units on eBay which have 5 outputs and in a nice form factor
which I like the look of, but I cannot find any internal pictures or circuit diagrams so I have no idea what's lurking inside!
Any advice on the above would be gratefully appreciated :-+
-
It is about 40 years since I was using broadcast spec 6-output video distribution amplifiers on a daily basis. Handling PAL composite video, RGB video and the 7 standard timing signals. (Horizontal drive, vertical drive, composite sync, mixed blanking, burst gate, PAL square wave and subcarrier.) They were certainly flat to over 5.5 MHz and probably nearly that to 10 MHz but I think they would have been significantly down at 30 MHz which is the first harmonic of your 10 MHz square wave so I expect what you get out will be nowhere near as square as your source.
IIRC the 6 pulse type signals were 2 Volt p-p into 75 ohms. Composite video and 4.43 MHz subcarrier were 1 Volt p-p into 75 ohms. Each of R,G and B were 0.7 Volts into 75 ohms unless G had syncs in which it was 1 Volt. To achieve those swings into 6 loads of 75 ohms requires quite high current in the output stage so they got quite warm! Modifying from 75 ohms to 50 ohms in and out means that at the same output levels 4 loads of 50 ohm will tax an amplifier the same as 6 of 75 ohm.
At that time they were 'all-discrete' designs. A single very low impedance output stage feeding the six 75 ohm output impedance resistors. I don't know if later designs still do that or use multiple ICs each feeding one output.
-
It is about 40 years since I was using broadcast spec 6-output video distribution amplifiers on a daily basis. Handling PAL composite video, RGB video and the 7 standard timing signals. (Horizontal drive, vertical drive, composite sync, mixed blanking, burst gate, PAL square wave and subcarrier.) They were certainly flat to over 5.5 MHz and probably nearly that to 10 MHz but I think they would have been significantly down at 30 MHz which is the first harmonic of your 10 MHz square wave so I expect what you get out will be nowhere near as square as your source.
IIRC the 6 pulse type signals were 2 Volt p-p into 75 ohms. Composite video and 4.43 MHz subcarrier were 1 Volt p-p into 75 ohms. Each of R,G and B were 0.7 Volts into 75 ohms unless G had syncs in which it was 1 Volt. To achieve those swings into 6 loads of 75 ohms requires quite high current in the output stage so they got quite warm! Modifying from 75 ohms to 50 ohms in and out means that at the same output levels 4 loads of 50 ohm will tax an amplifier the same as 6 of 75 ohm.
At that time they were 'all-discrete' designs. A single very low impedance output stage feeding the six 75 ohm output impedance resistors. I don't know if later designs still do that or use multiple ICs each feeding one output.
Thanks for your reply, it's been quite some time since I worked with analogue video and I forgot how much they had to squeeze in.
Though a couple of years back I dabbled with some amateur analogue TV on 1296MHz and acquired some TX/RX's in kit form that was all built incorrectly or partially built, my CRO was very useful at looking at those composite signals for fault finding. I got it all working correctly then shelved it, I thought digital is the way to go and wanted to try DATV on QO-100, but haven't got around to that either.
The models I've been looking at have either an opamp driving all the outputs, pairs of outputs or one for every output. I've not come across any with just discrete components as yet, but I have only been looking at "modern" products.
-
How about getting one of the many cheap 10MHz distribution amplifiers from Ebay? These are specifically designed for the purpose. I got the one designed by BG7TBL.
-
How about getting one of the many cheap 10MHz distribution amplifiers from Ebay? These are specifically designed for the purpose. I got the one designed by BG7TBL.
I did look at the BG7TBL distribution amplifiers, but by the time I import one the cost with VAT/Duty etc it works out almost the same price as buying another GPSDO.
How are you getting along with yours?
-
If 4 channels is enough, and if you can use a soldering iron, try building one of mine?
https://github.com/thinkfat/dist-amp
The MAX4022ESD+ I used is even available at Mouser, just the cost has gone up significantly in the last year. I bought them for about 4€ each, not they're 7€ per piece.
It's meant for sine wave input, but you can feed it with square wave, too, as long as the source impedance is 50\$\Omega\$. There's a low pass filter at the input and so the output will be sine wave, of course, but for synchronizing some test instruments it should be fine.
-
If 4 channels is enough, and if you can use a soldering iron, try building one of mine?
https://github.com/thinkfat/dist-amp
The MAX4022ESD+ I used is even available at Mouser, just the cost has gone up significantly in the last year. I bought them for about 4€ each, not they're 7€ per piece.
It's meant for sine wave input, but you can feed it with square wave, too, as long as the source impedance is 50\$\Omega\$. There's a low pass filter at the input and so the output will be sine wave, of course, but for synchronizing some test instruments it should be fine.
Cheers! Thanks for the link I'll have a good look at that, I do enjoy building something.
-
The AD8055 is perfect for this. Look
up feeltech fy6600 GitHub, under OXCO mods there is a schematic for a 10 MHz distributor with PLL. You only need the active filter, RF transformers and the AD8055. The design was contributed by one of the members here.
-
Cheers! Thanks for the link I'll have a good look at that, I do enjoy building something.
If you are OK building things, there are lots of options. Here is what I did:
- I found a generic video distribution amp somewhere -- flea market or eBay. Mine was a six-output model from Sigma Electronics, the "VDA-100A". That took care of the case, power supply, and BNC connectors.
- The original unit used a single high-power buffer stage, and six 75 ohm output resistors. I changed those to operate at 50 ohms, but the results were not satisfactory. So the existing buffer stage was removed from the pc board, leaving only the power supply.
- A promising replacement circuit was located on the time-nuts mailing list. That group is a great resource for problems like this. The following post describes a discrete amplifier with 10 channels. https://febo.com/pipermail/time-nuts_lists.febo.com/2014-November/071261.html (https://febo.com/pipermail/time-nuts_lists.febo.com/2014-November/071261.html). Here is the schematic: https://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20141126/88a31a74/attachment.gif (https://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20141126/88a31a74/attachment.gif)
- I cut it back to six channels, and built it up on some scrap copperclad which was then mounted onto the vacated space left by the original buffer stage.
Now I can drive a scope, counter, and some signal/function generators simultaneously and it all works as expected.
Some pictures of that project are attached.
Good luck whichever way you end up going.
-
I tried some video distribution amps. One worked well enough but the other did sync tip restoration and made a bit of a mess of the sinewave.
-
You can consider these from TAPR (Tucson Amateur Packet Radio):
- TADD-1 RF Distribution Amplifier https://tapr.org/product/tadd-1-rf-distribution-amplifier/ (https://tapr.org/product/tadd-1-rf-distribution-amplifier/)
- TADD-3 Pulse-Per-Second Distribution Amplifierhttps://tapr.org/product/tadd-3-pulse-per-second-distribution-amplifier/ (https://tapr.org/product/tadd-3-pulse-per-second-distribution-amplifier/)
-
How about getting one of the many cheap 10MHz distribution amplifiers from Ebay? These are specifically designed for the purpose. I got the one designed by BG7TBL.
I did look at the BG7TBL distribution amplifiers, but by the time I import one the cost with VAT/Duty etc it works out almost the same price as buying another GPSDO.
How are you getting along with yours?
I must day that I did upgrade the BNC connectors but other than that is works and it looks quite well designed. I'm not sure whether you can built something similar and not spend more on parts (and time).
-
Thank you all for the suggestions above, I've looked in to them all and quite like the TAPR products, but the distribution amplifier is on back order and will cost a small fortune to import it to the UK by the time VAT/Duty/Courier fees are added.
I will probably get some PCB's made and have a play around building some designs that have been linked to above.
At the moment I need something ready to go so I will order something like the BG7TBL unit.
How about getting one of the many cheap 10MHz distribution amplifiers from Ebay? These are specifically designed for the purpose. I got the one designed by BG7TBL.
I did look at the BG7TBL distribution amplifiers, but by the time I import one the cost with VAT/Duty etc it works out almost the same price as buying another GPSDO.
How are you getting along with yours?
I must day that I did upgrade the BNC connectors but other than that is works and it looks quite well designed. I'm not sure whether you can built something similar and not spend more on parts (and time).
I'm sorely tempted to order one, do you feed yours with a squarewave or sinewave and do you know if it the output is the same as what is inputted or is there a filter?
The only place I can find these is on Aliexpress and it seems to be the same price with or without the OCXO, looking at the PCB I'm assuming the OCXO is not disciplined by the GPSDO and is only active when there is no external 10MHz reference? I wondered if there is any disadvantage to getting the one with a OCXO as that may be handy for backup.
I found another 10MHz OCXO/Distribution unit on eBay from a seller called Pincei_de who are based in China, this is a similar device but has a different layout to the BG7TBL externally. Though I cannot find any information about this anywhere online, this forum or Aliexpress, I have attached a picture below.
-
Thank you all for the suggestions above, I've looked in to them all and quite like the TAPR products, but the distribution amplifier is on back order and will cost a small fortune to import it to the UK by the time VAT/Duty/Courier fees are added.
I will probably get some PCB's made and have a play around building some designs that have been linked to above.
At the moment I need something ready to go so I will order something like the BG7TBL unit.
How about getting one of the many cheap 10MHz distribution amplifiers from Ebay? These are specifically designed for the purpose. I got the one designed by BG7TBL.
I did look at the BG7TBL distribution amplifiers, but by the time I import one the cost with VAT/Duty etc it works out almost the same price as buying another GPSDO.
How are you getting along with yours?
I must day that I did upgrade the BNC connectors but other than that is works and it looks quite well designed. I'm not sure whether you can built something similar and not spend more on parts (and time).
I'm sorely tempted to order one, do you feed yours with a squarewave or sinewave and do you know if it the output is the same as what is inputted or is there a filter?
The only place I can find these is on Aliexpress and it seems to be the same price with or without the OCXO, looking at the PCB I'm assuming the OCXO is not disciplined by the GPSDO and is only active when there is no external 10MHz reference? I wondered if there is any disadvantage to getting the one with a OCXO as that may be handy for backup.
I found another 10MHz OCXO/Distribution unit on eBay from a seller called Pincei_de who are based in China, this is a similar device but has a different layout to the BG7TBL externally. Though I cannot find any information about this anywhere online, this forum or Aliexpress, I have attached a picture below.
I have this above mentioned black gear,
as always, you will never know what OXCO's are build in.
1. So you have to ask for a picture...
2. On my gear, a log amplier was mounted 180° false
3. usual, first converted to digital
4. than x times digital fan out drivers
5. each than by BP, for 10MHz to get the sine back
at the end, the PN of the OXCO is than for shure any lower..
-
I have this above mentioned black gear,
as always, you will never know what OXCO's are build in.
1. So you have to ask for a picture...
2. On my gear, a log amplier was mounted 180° false
3. usual, first converted to digital
4. than x times digital fan out drivers
5. each than by BP, for 10MHz to get the sine back
at the end, the PN of the OXCO is than for shure any lower..
Many thanks for the information this Black unit, have you posted on here about it with a teardown or anything?
Are you happy with it or are there still issues?
-
How about getting one of the many cheap 10MHz distribution amplifiers from Ebay? These are specifically designed for the purpose. I got the one designed by BG7TBL.
I did look at the BG7TBL distribution amplifiers, but by the time I import one the cost with VAT/Duty etc it works out almost the same price as buying another GPSDO.
How are you getting along with yours?
I must day that I did upgrade the BNC connectors but other than that is works and it looks quite well designed. I'm not sure whether you can built something similar and not spend more on parts (and time).
I'm sorely tempted to order one, do you feed yours with a squarewave or sinewave and do you know if it the output is the same as what is inputted or is there a filter?
I feed it with a sinewave and the output is also sinewave. Sinewave works excellent. For some projects I use a Cesium clock as a reference and those setups are using sine wave. At 10MHz the zero crossings of a sine wave are very steep already.
-
Many thanks for the information this Black unit, have you posted on here about it with a teardown or anything?
Are you happy with it or are there still issues?
What I do not like in terms of PN contribution is the Log amplifier and over all digital low speed IC 8x distribution.
I modified it to get on some outputs pure digital signals than sine.
I am very interested to have low PN contribution, so I am special on that matter. Usual the given OXCO as other OXCO module combos, will no any longer have better TAPU brand. Now given usually with unknown PN figures.
So be prepared of :palm:
-
Some tear down of the black and the BG7TBL....
-
Some tear down of the black and the BG7TBL....
Wow they are using discrete components in the black one.
The BG7TBL does look a better design.