Date   
Re: AD9850/51 module pc boards

William R Maxwell
 

Same here, Clifford

On 1/10/2018 5:31 PM, Clifford Heath wrote:
On 1 Oct 2018, at 4:42 pm, Don Vosper via Groups.Io <donald.vosper=btinternet.com@groups.io> wrote:
I see on the Farnell site in the UK that one off 9851, 9850 and the useful AD8037 are £25.00, £23.00 and £10.00 each respectively.
My last 80 or so purchases have been through AliExpress.
Typically you get parts for under the 10,000 unit price.
I think I've twice had a situation where a completely
wrong thing was shipped, and I didn't pay for those after
disputing them.

All other parts have been up to normal spec, and were
shipped free of charge. Most of the world's electronics is
made in China. There are a lot of "partial reels" of perfectly
good parts being sold off through the thousands of vendors
on AliExpress. I'm really struggling to see any downside.

Clifford Heath.


Re: AD9850/51 module pc boards

Glenn
 

Hi all, as it happens I had been working on a replacement pcb for the "type 2" board.

I made one some months back and had some produced by a Chinese supply house, but it had an error in the serial port section. I have reworked it though to fix that error and also to add the 3v3 regulator for the oscillator. Its 41.5 x 30mm in size, pretty close to the originals. It also was reworked for 5o ohm output and the LPF changed accordingly with a cut-off of about 44MHz. The LPF can be built using smd inductors or you can wind your own on toroids. The output of the AD9850 itself uses a trifilar wound transformer on a BN43-2402 core. (I did not footprint a commercial transformer as Terry has done)

I have not though had this board pro made. (I fixed the small error on the 1st version, running a couple of wires & track cuts.) Currently the Gerbers result in 6 boards on a panel with V-grooving between boards. Sch and Gerber pic of the panel attached.

glenn
vk3pe

Re: AD9850/51 module pc boards

Terry VK5TM
 

When the Pocket Money Fairy deigns to leave a few shekels under the pillow, I will do a run of 10 pcb's to check that there are no errors.

Probably take in the region of 4 weeks by the time the pcb's arrive (I don't do DHL -> rip-off merchants if you live outside a major city).

Re: AD9850/51 module pc boards

William R Maxwell
 

I guess at our age, a gift from the Pocket Money Fairy is more likely than the Tooth Fairy, Terry.

Know what you mean about DHL  when you live outside the metropolis. I'm 30 kms south of Launcesaton, on the main north-south highway. I'm sometimes lucky and a local courier delivers when driving past the door but more likely than not, its a trip into the depot.

Still, at the size of these boards there are currently some incredible offers from Chinee fab houses. I had some made by JLCPCB.om last week, just before they shut for a 3 day holiday. Mine were slightly too big to qualify but the claasic 100x100mm size is again on offer at $2 USD for 10, plus shipping. At that price even DHL shipping seems almost bearable, although they offer cheaper alternatives too.

Bill VK7MX

On 3/10/2018 3:05 PM, Terry VK5TM wrote:
When the Pocket Money Fairy deigns to leave a few shekels under the pillow, I will do a run of 10 pcb's to check that there are no errors.

Probably take in the region of 4 weeks by the time the pcb's arrive (I don't do DHL -> rip-off merchants if you live outside a major city).


Re: AD9850/51 module pc boards

Terry VK5TM
 

Hi Bill.

My problem is, because I am supposedly in the most remote of locations according to them (lets see them deliver to Birdsville), they doubled the shipping cost then added a remote fee on top of that.

Used them once, $68 shipping fee on a $30 order - not doing that again.

PS, I sent a private email to your bigpond address.

Re: AD9850/51 module pc boards

William R Maxwell
 

Had that experience too, Terry. Actually, DHL agents in China seem to be particularly adept at deeming parts of Australia as remote, with absolutely no local knowledge, I suspect. I'm almost exactly 8 nautical miles from Launceston Airport, where most incoming airfreight arrives in Tassie but no, I'm remote too in some Chinese DHL assessments!

Thanks for the private email. I'll take a look at it now.

Bill VK7MX

On 3/10/2018 4:51 PM, Terry VK5TM wrote:
Hi Bill.

My problem is, because I am supposedly in the most remote of locations according to them (lets see them deliver to Birdsville), they doubled the shipping cost then added a remote fee on top of that.

Used them once, $68 shipping fee on a $30 order - not doing that again.

PS, I sent a private email to your bigpond address.


Re: AD9850/51 module pc boards

VK5ZIF
 

Hi Terry,

How many shekels do you need and how do you prefer it sent?

I'll take one please.

73

Iain VK5ZIF

On 3/10/2018 2:35 PM, Terry VK5TM wrote:
When the Pocket Money Fairy deigns to leave a few shekels under the pillow, I will do a run of 10 pcb's to check that there are no errors.

Probably take in the region of 4 weeks by the time the pcb's arrive (I don't do DHL -> rip-off merchants if you live outside a major city).

Type 2 Build QRV

Curt
 

okay first the boring status: supply chain (me) has ordered a Uno, some headers and a type 2 AD9850 board from the far east - so it should land here someday and likely soon enough. I mounted the SMT passives to the detector board - wow this is fun in spite of not having good eyesight. I have the ICs in hand and/or on order.

I did figure out which side was ground for the board and that tantalum.

I noticed the AD820 I have are surface mount. I am so tempted to place one on one of the pads and gently wire the few other connections needed (I have done this before. If this isn't sane - tell me - I have PDIPs for two alternate parts with LT in the number on route).

Looks like I have most everything else on hand -- well except for those resettable fuses that I am not familiar with. Those things do look cool - time to see if I can get an assortment maybe with those two values.

Good to be started - that was the hardest part thus far. Thanks to the designers (yes I have fond memories of dialog with Jerry and I have two of his paddles in use here). Glad a few folk are still on list when I have issues.

73 Curt

Re: Type 2 Build QRV

William Heller
 

If you can afford it digikey probably has a kit of the resettable fuses try
littlefuse on there search box.

On Wed, Oct 3, 2018, 6:44 PM Curt via Groups.Io <wb8yyy=yahoo.com@groups.io>
wrote:

okay first the boring status: supply chain (me) has ordered a Uno, some
headers and a type 2 AD9850 board from the far east - so it should land
here someday and likely soon enough. I mounted the SMT passives to the
detector board - wow this is fun in spite of not having good eyesight. I
have the ICs in hand and/or on order.

I did figure out which side was ground for the board and that tantalum.

I noticed the AD820 I have are surface mount. I am so tempted to place
one on one of the pads and gently wire the few other connections needed (I
have done this before. If this isn't sane - tell me - I have PDIPs for two
alternate parts with LT in the number on route).

Looks like I have most everything else on hand -- well except for those
resettable fuses that I am not familiar with. Those things do look cool -
time to see if I can get an assortment maybe with those two values.

Good to be started - that was the hardest part thus far. Thanks to the
designers (yes I have fond memories of dialog with Jerry and I have two of
his paddles in use here). Glad a few folk are still on list when I have
issues.

73 Curt





Re: AD9850/51 module pc boards

Terry VK5TM
 

Iain, private message sent to your bigpond address.

Terry VK5TM

Finished at last

K5ESS
 

Finally completed my PHSNA (after three years). Built it in a re-purposed HP-435 chassis. Took advantage of the 435's 50MHz, 1mW power reference and included it. Also included a second channel. Guess I'll have to write my own software to take advantage of the second channel. Originally thought I'd have room for an internal battery but turned out not to be the case, so for now the battery voltage scale is of no use unless and until I add provisions for an external battery.
Mike
K5ESS

Re: Finished at last

sohosources
 

Awesome! Thanks for sharing this.

--Kirk, NT0Z
Rochester, MN

------ Original Message ------
From: "K5ESS" <k5ess.nothdurft@...>
To: PHSNA@groups.io
Sent: 10/24/2018 9:31:09 PM
Subject: [PHSNA] Finished at last

Finally completed my PHSNA (after three years). Built it in a re-purposed HP-435 chassis. Took advantage of the 435's 50MHz, 1mW power reference and included it. Also included a second channel.
Guess I'll have to write my own software to take advantage of the second channel. Originally thought I'd have room for an internal battery but turned out not to be the case, so for now the battery voltage scale is of no use unless and until I add provisions for an external battery.
Mike
K5ESS


Re: Finished at last

Steven Dick
 

Wow. Just beautiful. very nice packaging!!

-----Original Message-----
From: K5ESS
Sent: Wednesday, October 24, 2018 10:31 PM
To: PHSNA@groups.io
Subject: [PHSNA] Finished at last

Finally completed my PHSNA (after three years). Built it in a re-purposed HP-435 chassis. Took advantage of the 435's 50MHz, 1mW power reference and included it. Also included a second channel. Guess I'll have to write my own software to take advantage of the second channel. Originally thought I'd have room for an internal battery but turned out not to be the case, so for now the battery voltage scale is of no use unless and until I add provisions for an external battery.
Mike
K5ESS

Re: AD9850/51 module pc boards

Terry VK5TM
 

The prototype versions of the boards arrived yesterday.

I'm sending a couple out to two members here and will post some more info once they have been built and debugged.

FS: RF Power Meter Board

Larry Naumann <n0sa@...>
 

I was cleaning off my work bench and found this item.
It is the board for the W7ZOI/W7PUA power meter.
EAM Solutions Nov, 10, 2015 rev 2.00 sch.
It is all built and fully populated but never checked out.
Should work just fine.
Make me a reasonable offer with shipping and it is yours.
please contact direct to me.
n0sa@...
73
Larry
n0sa

Testing the Measurement Rcvr

Nigel Maund
 

Dear PHSNA testing community;

I am finally building the Measurement Receiver.

I wanted to ask advice on how to test/tune it as I build it in sections?

Here are my thoughts:

1) using the PHSNA, sweep the Xtal filter alone, and adjust the trim capacitors to get a flat response curve
2) add the 2n2222 Amplifier, to then see the sweep with a 20 dB gain

Next steps - how would I make sure that the Bridge Tee Diplexers are properly tuned?

Is it simple enough to wire up all the components and go for broke.
Or, is there a way to perhaps add in the Diplexer circuit and sweep it using a 3.2768 MHz (spare crystal) oscillator,. If I were to use a return loss bridge would this then give me some interesting way to use all the test components and actually see the Diplexer working?

I am thinking of how best to test each of the two boards separately, using the PHSNA and Power Meter, before connecting them together.

One final question: the schematic shows only one power connection circuitry (200 mA resettable fuse, shotkey diode) but both boards have PCB locations for this circuitry. Is it acceptable to simply wire one board with this circuitry, and jumper the 12 volts supply to the 2nd board. Or, do both boards need a 200 mA resettable fuse?

Thanks,
Nigel, Va2NM

Re: Testing the Measurement Rcvr

Nick Kennedy
 

It's been 4+ years since I built mine, but I have some notes, fortunately.
Your idea of testing in sections is certainly a good one. I had to do a lot
of troubleshooting, although not generally due to the design or concept but
due to having a couple of bad mixers.

Also, I made a lot of changes so I had to be sure I'd done them correctly.
I used some 5.528 MHz crystals I had on hand so that changed the operating
frequency, the diplexer, the L-matches and so on.

Shooting through the crystal filter and its L-matchers before connecting to
the rest of the world is definitely a good idea. And since you have that
nice PHSNA instrument sitting there, why not?

I like to measure my critical component values pretty closely and also
simulate the circuits in LTspice, so sometimes I don't actually feel the
need to adjust trimmers. My notes say I pre-set my L-match trimmers to the
required value. But I could see plotting the filter curve and playing with
those trimmers until you got a nice flat response curve might be a good
thing.

I substituted an ERA-1SM for the ERA-3+ which changed other component
values a bit. It's easy enough to verify the gain of that stage acting
alone before installing the mixer.

For whatever reason(s), I didn't feel the need for so much oomph out of the
output amplifier so I changed the biasing to reduce standing current to
about 10 mA and didn't need a heat sink.

I did take a look through the diplexer before connecting it to the driving
and driven stages. I had already modeled it on LTspice and seen that, by
design, you don't get sharp peaks but you do get a constant load over a
wide range of frequencies. Which I guess is its function.

So getting back to testing in stages. Yes, especially on through-hole board
projects, it's a great idea. Sections can be hard to break apart after the
fact. And as I guess I've suggested already, you could easily build and
test separately the MMIC amplifier, the diplexer, the crystal filter plus
matchers and the output amp. Or combine the output amp with the filter.
This can be accomplished by holding off on installing the mixer until last,
testing the boards separately, and maybe keeping C15 out until you've
tested the filter and output amplifier separately.

On the fuse and blocking diode - your proposal to share them seems
reasonable. I don't remember exactly what I did there.

73 & good luck with the project,

Nick, WA5BDU

On Tue, Jan 1, 2019 at 6:40 PM Nigel Maund <maund.n@...> wrote:

Dear PHSNA testing community;

I am finally building the Measurement Receiver.

I wanted to ask advice on how to test/tune it as I build it in sections?

Here are my thoughts:

1) using the PHSNA, sweep the Xtal filter alone, and adjust the trim
capacitors to get a flat response curve
2) add the 2n2222 Amplifier, to then see the sweep with a 20 dB gain

Next steps - how would I make sure that the Bridge Tee Diplexers are
properly tuned?

Is it simple enough to wire up all the components and go for broke.
Or, is there a way to perhaps add in the Diplexer circuit and sweep it
using a 3.2768 MHz (spare crystal) oscillator,. If I were to use a return
loss bridge would this then give me some interesting way to use all the
test components and actually see the Diplexer working?

I am thinking of how best to test each of the two boards separately, using
the PHSNA and Power Meter, before connecting them together.

One final question: the schematic shows only one power connection
circuitry (200 mA resettable fuse, shotkey diode) but both boards have PCB
locations for this circuitry. Is it acceptable to simply wire one board
with this circuitry, and jumper the 12 volts supply to the 2nd board. Or,
do both boards need a 200 mA resettable fuse?

Thanks,
Nigel, Va2NM



Re: Testing the Measurement Rcvr

Nigel Maund
 

Thanks for your feedback Nick,

I am intrigued as to why you may have changed to the 5.528 MHz frequency of operation?

And, also the decisions to go with lower amplifier output.

In researching other types of analyzers, I came across Ashar Farhan’s “Specan” design.

http://hfsignals.blogspot.com/p/specan-reboot-of-w7zoi.html
https://drive.google.com/file/d/0BzRNYeu10K6DdE1CSDVfZUdTdzg/view

He used a 12 MHz filter.
It occurs to me, that I could build the front stages of his design, and “tack them on” as front stages to the Measurement Receiver.

Let me start by getting the Measurement Receiver built (it has been, like 4 years on the back burner!).

Thanks and interested to hear from others that have implemented the Measurement Receiver.

Rgds,
Nigel
VA2NM

On Jan 2, 2019, at 8:31 PM, Nick Kennedy <kennnick@...> wrote:

It's been 4+ years since I built mine, but I have some notes, fortunately.
Your idea of testing in sections is certainly a good one. I had to do a lot
of troubleshooting, although not generally due to the design or concept but
due to having a couple of bad mixers.

Also, I made a lot of changes so I had to be sure I'd done them correctly.
I used some 5.528 MHz crystals I had on hand so that changed the operating
frequency, the diplexer, the L-matches and so on.

Shooting through the crystal filter and its L-matchers before connecting to
the rest of the world is definitely a good idea. And since you have that
nice PHSNA instrument sitting there, why not?

I like to measure my critical component values pretty closely and also
simulate the circuits in LTspice, so sometimes I don't actually feel the
need to adjust trimmers. My notes say I pre-set my L-match trimmers to the
required value. But I could see plotting the filter curve and playing with
those trimmers until you got a nice flat response curve might be a good
thing.

I substituted an ERA-1SM for the ERA-3+ which changed other component
values a bit. It's easy enough to verify the gain of that stage acting
alone before installing the mixer.

For whatever reason(s), I didn't feel the need for so much oomph out of the
output amplifier so I changed the biasing to reduce standing current to
about 10 mA and didn't need a heat sink.

I did take a look through the diplexer before connecting it to the driving
and driven stages. I had already modeled it on LTspice and seen that, by
design, you don't get sharp peaks but you do get a constant load over a
wide range of frequencies. Which I guess is its function.

So getting back to testing in stages. Yes, especially on through-hole board
projects, it's a great idea. Sections can be hard to break apart after the
fact. And as I guess I've suggested already, you could easily build and
test separately the MMIC amplifier, the diplexer, the crystal filter plus
matchers and the output amp. Or combine the output amp with the filter.
This can be accomplished by holding off on installing the mixer until last,
testing the boards separately, and maybe keeping C15 out until you've
tested the filter and output amplifier separately.

On the fuse and blocking diode - your proposal to share them seems
reasonable. I don't remember exactly what I did there.

73 & good luck with the project,

Nick, WA5BDU


On Tue, Jan 1, 2019 at 6:40 PM Nigel Maund <maund.n@... <mailto:maund.n@...>> wrote:

Dear PHSNA testing community;

I am finally building the Measurement Receiver.

I wanted to ask advice on how to test/tune it as I build it in sections?

Here are my thoughts:

1) using the PHSNA, sweep the Xtal filter alone, and adjust the trim
capacitors to get a flat response curve
2) add the 2n2222 Amplifier, to then see the sweep with a 20 dB gain

Next steps - how would I make sure that the Bridge Tee Diplexers are
properly tuned?

Is it simple enough to wire up all the components and go for broke.
Or, is there a way to perhaps add in the Diplexer circuit and sweep it
using a 3.2768 MHz (spare crystal) oscillator,. If I were to use a return
loss bridge would this then give me some interesting way to use all the
test components and actually see the Diplexer working?

I am thinking of how best to test each of the two boards separately, using
the PHSNA and Power Meter, before connecting them together.

One final question: the schematic shows only one power connection
circuitry (200 mA resettable fuse, shotkey diode) but both boards have PCB
locations for this circuitry. Is it acceptable to simply wire one board
with this circuitry, and jumper the 12 volts supply to the 2nd board. Or,
do both boards need a 200 mA resettable fuse?

Thanks,
Nigel, Va2NM



For Sale: Type-2 PHSNA

N5IB
 

Doing a bit of bench clearing...

For sale is an assembled, programmed, and working Type-2 PHSNA, including
the Type II DDS module, with its on-board LPF removed
an Arduino UNO
both of the above mounted on an assembled PHSNA motherboard
an 8307 RF power detector board with front end shield
all of the above mounted "breadboard style" as shown in the attached photo

This is actually the prototype build of the Type-2 PHSNA
$50 plus shipping costs.
Unfortunately this will have to be confined to USA shipping, as the international rate will be outrageous,
though Canada might not be unreasonably costly

if interested, reply off-list to n5ib (at) juno (dot) com

BTW - I'm thinking of parting with my TenTec/TAPR vector network analyzer - if you're interested, e-mail at the above address and we can discuss its "special needs"

Jim

Re: Testing the Measurement Rcvr

Nick Kennedy
 

I went to the 5.528 MHz IF solely in the spirit of trying to use parts that
are in my junk box. It certainly has no advantage over the frequency chosen
by Jerry H.

Some of the other stuff I did is hard to explain now. Possibly I wanted to
not use a heat sink on the output amplifier. And I felt that the PHSNA has
plenty of dynamic range. It's definitely a good design as-is though.

Yes, Farhan's SPECAN is very interesting. Who wouldn't love to have a
spectrum analyzer? And he claims it's easy enough to build and not
expensive. Maybe that one should go on the stack.

73-

Nick, WA5BDU

On Thu, Jan 3, 2019 at 11:47 AM Nigel Maund <maund.n@...> wrote:

Thanks for your feedback Nick,

I am intrigued as to why you may have changed to the 5.528 MHz frequency
of operation?

And, also the decisions to go with lower amplifier output.

In researching other types of analyzers, I came across Ashar Farhan’s
“Specan” design.

http://hfsignals.blogspot.com/p/specan-reboot-of-w7zoi.html
https://drive.google.com/file/d/0BzRNYeu10K6DdE1CSDVfZUdTdzg/view

He used a 12 MHz filter.
It occurs to me, that I could build the front stages of his design, and
“tack them on” as front stages to the Measurement Receiver.

Let me start by getting the Measurement Receiver built (it has been, like
4 years on the back burner!).

Thanks and interested to hear from others that have implemented the
Measurement Receiver.

Rgds,
Nigel
VA2NM

On Jan 2, 2019, at 8:31 PM, Nick Kennedy <kennnick@...> wrote:

It's been 4+ years since I built mine, but I have some notes,
fortunately.
Your idea of testing in sections is certainly a good one. I had to do a
lot
of troubleshooting, although not generally due to the design or concept
but
due to having a couple of bad mixers.

Also, I made a lot of changes so I had to be sure I'd done them
correctly.
I used some 5.528 MHz crystals I had on hand so that changed the
operating
frequency, the diplexer, the L-matches and so on.

Shooting through the crystal filter and its L-matchers before connecting
to
the rest of the world is definitely a good idea. And since you have that
nice PHSNA instrument sitting there, why not?

I like to measure my critical component values pretty closely and also
simulate the circuits in LTspice, so sometimes I don't actually feel the
need to adjust trimmers. My notes say I pre-set my L-match trimmers to
the
required value. But I could see plotting the filter curve and playing
with
those trimmers until you got a nice flat response curve might be a good
thing.

I substituted an ERA-1SM for the ERA-3+ which changed other component
values a bit. It's easy enough to verify the gain of that stage acting
alone before installing the mixer.

For whatever reason(s), I didn't feel the need for so much oomph out of
the
output amplifier so I changed the biasing to reduce standing current to
about 10 mA and didn't need a heat sink.

I did take a look through the diplexer before connecting it to the
driving
and driven stages. I had already modeled it on LTspice and seen that, by
design, you don't get sharp peaks but you do get a constant load over a
wide range of frequencies. Which I guess is its function.

So getting back to testing in stages. Yes, especially on through-hole
board
projects, it's a great idea. Sections can be hard to break apart after
the
fact. And as I guess I've suggested already, you could easily build and
test separately the MMIC amplifier, the diplexer, the crystal filter plus
matchers and the output amp. Or combine the output amp with the filter.
This can be accomplished by holding off on installing the mixer until
last,
testing the boards separately, and maybe keeping C15 out until you've
tested the filter and output amplifier separately.

On the fuse and blocking diode - your proposal to share them seems
reasonable. I don't remember exactly what I did there.

73 & good luck with the project,

Nick, WA5BDU


On Tue, Jan 1, 2019 at 6:40 PM Nigel Maund <maund.n@... <mailto:
maund.n@...>> wrote:

Dear PHSNA testing community;

I am finally building the Measurement Receiver.

I wanted to ask advice on how to test/tune it as I build it in sections?

Here are my thoughts:

1) using the PHSNA, sweep the Xtal filter alone, and adjust the trim
capacitors to get a flat response curve
2) add the 2n2222 Amplifier, to then see the sweep with a 20 dB gain

Next steps - how would I make sure that the Bridge Tee Diplexers are
properly tuned?

Is it simple enough to wire up all the components and go for broke.
Or, is there a way to perhaps add in the Diplexer circuit and sweep it
using a 3.2768 MHz (spare crystal) oscillator,. If I were to use a
return
loss bridge would this then give me some interesting way to use all the
test components and actually see the Diplexer working?

I am thinking of how best to test each of the two boards separately,
using
the PHSNA and Power Meter, before connecting them together.

One final question: the schematic shows only one power connection
circuitry (200 mA resettable fuse, shotkey diode) but both boards have
PCB
locations for this circuitry. Is it acceptable to simply wire one board
with this circuitry, and jumper the 12 volts supply to the 2nd board.
Or,
do both boards need a 200 mA resettable fuse?

Thanks,
Nigel, Va2NM