Ascel AE20401 2Hz to 5.8GHz frequency counter + RF power meter kit

[Fraser]

Some background:

I own several nice quality frequency counters from the likes of Advantest, Anritsu, Tabor , Racal and Philips. They all have one issue that I needed to address. Non were capable of measuring frequencies above around 1.5GHz. My Advantest SA has a frequency counter capable of 3GHz but that is still not good enough for some modern technology working at 5.8Ghz.

I built pre-scalers based on Hittite chips but these are cumbersome to use as you have to calculate their output as none are /10 or /100 types. I considered building a /10 10Ghz prescaler but it was costly and I never got around to it.

I had seen the ASCEL frequency counter and power meter kit some time ago but hadn't really given it much thought since. The recent improvement in £/Euro exchange rate has made it more appealing though and I revisited the product.

What is the ASCEL AE20401 ?

It is a kit of parts to make a unit that can measure frequency and RF power. The kit comes in basic form or with additional options that the user selects at time of purchase or as a retrofit later. The options are a nice case, USB PC connectivity,  TCXO timebase module, RF power measurement module (up to 500MHz) and most important to me... a 5.8GHz prescaler module. Each module is availble to buy separately but they are designed for teh AE20401 so use with other brands of equipment would need modifications.

I worked out the total cost of the AE20401 kit with ALL options included to be £108. That seems a reasonable price when considering the included parst and high quality of the kit.

When built, you have a compact unit that can provide accurate frequency and RF power measurement. The specifications are frequency measurement from 2Hz to 5.8GHz, and RF Power measurement from DC to 500MHz. The unit may be interfaced with a PC vis USB providing data collection and graphing etc. Software is provided.

My kit arrived today. I attach pictures of the content. First impressions are that this is a well designed kit that uses quality components. All SMD components are pre-fitted with only through hole to be soldered into place by the builder. The OCXO, RF Power and 5.8GHz prescaler modules are all pre-built and just connect to the pin headers on the PCB.

PCB and case quality is excellent. This looks like it will be a very neat unit once built. It is also small enough to be considered hand portable and would fit into a mobile kit easily.

The manual may be found here:

http://www.ascel-electronic.de/blog/detail/sCategory/10/blogArticle/9

The kit may be viewed here:

http://www.ebay.co.uk/itm/AE20401-5-8-GHz-Frequency-Counter-RF-Power-Meter-Pulse-Counter-Kit-with-USB-/261582387430?pt=LH_DefaultDomain_0&var=&hash=item3ce78678e6

But I got it chepaer by buying in Euros 

http://www.ebay.co.uk/itm/AE20401-5-8-GHz-Frequenzzahler-Power-Meter-Impulszahler-Komplett-Bausatz-/261581508905?pt=LH_DefaultDomain_77&var=&hash=item3ce7791129

Instructions came in English so do not worry that the auction is in German !

I also purchased a spare 5.8GHz prescaler module as a spare of possible use with another frequency counter.

No idea when I will get this built but thought readers may be interested in seeing the kit.

Pictures attached.

P.S...... I went for the Black on Green/Yellow display as I am starting to really hate White on Blue LCD displays. The LCD module id a standard type so the owner can change it in the future if desired.

[Fraser]

The sting in this tail of this story is that after ordering the AE20401, I saw an advert on this forum for a lovely EIP 25B 20GHz selective microwave frequency counter that includes RF power measurement above 1GHz .... perfect for my needs 

Well to cut a long story short, that 20GHz frequency counter is now on its way to me. The AE20401 has effectively been made redundant by the EIP but I am sure I will find uses for it, and it is very 'cute' 

Aurora

[Lightages]

That looks like a nice kit. I am starting to see this brand on a number of people's benches in some youtube videos. The function generator looks like it might be OK too.

[Fraser]

@Lightages,

It certainly looks nicely produced. It is also good to see that ASCEL are happy to provide spares such as a replacement microprocessor or LCD for around 4 Euros !

http://www.ascel-electronic.de/components/

I took a look at the 10Mhz function generator as well. I do not need another one, but I was interested to see what DDS they are using. I see that it is the AD9833 in this unit. Looking at the limited filtering it will likely suffer the same spurii issues as other similar DDS FG's, but it still appears reasonable value for money at around 58 Euros (with USB connectivity).

I know nothing of ASCEL as a company but they appear to have a limited product line containing the frequency counter, LC meter and function generator. I am wondering if they are a small company wishing to make a name for themselves before expanding their product range. A bit like a German version of PEAK Electronics who make the DCA75, ESR70 etc. If so, I wish them every success 

Aurora

[Fraser]

I must raise one negative comment regarding the design of this unit...... what were they thinking when they chose to use an type 'A' (Host) USB connector on a USB type 'B' (non-host) appliance !

It was a decision likely driven by connector shape and dimensions as the front panel is a little tight for free space. A Mini type B USB connector would have been a better choice however.

A minor issue and I note that they used the normal type 'B' USB connector on the LC meter.

Aurora

[Fraser]

There is not that much work needed to build this simple little kit, so I may well find time this week to construct it. I will be interested to see how well it performs.

Watch this space 

Aurora

[G0HZU]

Considering it is 'new' then it does appear to offer some useful functionality for the kit price.

I'm not too impressed with the front panel silk screen though. It looks a bit cluttered and clumsy. But one bonus is that the fonts are big and will be easy to read for many users even in poor lighting.

Some users may wish it had an external reference option but if it can deliver 1ppm reliably between calibrations then it should be fine for many.
The power meter looks to be based on the classic AD8307 logamp so this is probably why they claim performance up to 500MHz but there will probably be some cal factors to include for best performance across the whole 500MHz range.

I used to lust after the EIP counters in the 1980s especially the source locking 545 model. We had several of these at work. However, the power meter option in your EIP 25 is probably going to be a bit dodgy in terms of accuracy and uncertainty as I think these use a Schottky diode as a detector and the calibration process is very complicated with lots of adjustment points to correct for its inherently poor flatness across frequency and level. So the cal lookup tables may be less than ideal unless it's had a recent service. Also the input match isn't going to be as good as a regular microwave power meter so maybe fit an attenuator here to minimise uncertainty.

[dadler]

It's crazy how $100 DIY kits these days come with software that seems 100x better designed and more useful than any of the Rigol (et al) software. I have no explicit use for one of these (or the function gen) but I am tempted just to try the hardware+software together.

[Fraser]

@dadler,

It would appear that some major manufacturers employ the 'B' team when it comes to getting PC software written, The classic case is Hantek equipment.....nice hardware and truly appalling software.

I have also just purchased a used DG8SAQ VNWA3 VNA that is pre built by SDR-Kits. It used to be a kit but is now only available in built form. The software that drives it is very impressive and is the result of many iterations and improvements by the author.

There is a lot of talent in the user community and this is amply demonstrated in this forums threads. I believe that some truly excellent software can be created by 'amateurs' (no offence intended by the term) especially when they receive guidance and feedback from those with more experience on forums such as this.

Some manufacturers take the 'it will do' approach to software, whereas many user communities and small companies take the far more positive view of, "well it works, but how can I further improve it ?"

I am very impressed with the quality of work that small dedicated companies and 'amateurs' can produce, and continue to improve.

Aurora

[nowlan]

I take your word these are good products for the price.
Is the frequency using the Euro comma/decimal place, or is it really 78 hz?
Agree that the buttons seem to be in poor position, once cables plugged in.

[Fraser]

@nowlan,

Do not take my word for anything....I have not built it yet 

Make your own decisions based upon the pictures I took and the schematics that appear in the downloadable manual. My opinion of quality relates to the parts chosen and the PCB quality. The separate Power and 5.8GHz module appear very well made and 'professional'. IMHO £108 for a 5.8GHz capable frequency counter is very reasonable. The fact that you get full schematics and a design description in the manual enables customisation as well.

Once I have this thing built I will be in a position to comment upon ease of use and ergonomics.

Regarding the case..... the designer appears to have had two key objectives, build the unit on a single PCB that sits behind the front panel, and to make it fit a certain size of case. The result is a nice compact PCB, but also a crowded front panel. We shall have to see whether such compact dimensions become a hindrance. It certainly makes the unit portable and it can run off of a battery supply 

The master oscillator is 20MHz so adding an external frequency standard input would involve making a frequency doubler from the standard 10MHz reference. Not a complex task if desired and just a simple SPDT switch used to select which reference is used.

I was going to build the unit today but the pain in my arm joints (M.E) is too great to do a decent soldering job so I have postponed the build till I feel better. Hopefully this week.

Aurora

[Fraser]

@G0HZU,

Regarding the EIP 25B........ it is not a bolometer based power meter as you say. The unit behaves more like a spectrum analyser above 1GHz and so measures peak power in a similar manner in a detector stage. Accuracy will be adequate for my needs though.

For those unaware, it was very hard to produce a frequency counter that could operate above 1GHz without using a pre-scaler and they used to be limited to around 2GHz. This is why many of the used professional frequency counters topped out at less than 2GHz (as is the case with my other units). Ties have changed and much higher frequency pre-scalers are now available from the likes of Hittite. Microwave frequency counters are unusual beasts however and not just a pre-scaled conventional counter....to learn more read on......

The EIP 25B microwave frequency counter is effectively three devices in one case. 1. a conventional frequency counter, 2. a YIG tuned heterodyne down-converter and 3. an RF power meter (above 1GHz)

I shall not go into the conventional counter theory as that is easily found on the NET.
The interesting bit (for me) is the >1GHz heterodyne down-converter stage for use up to 20GHz. It consists of an input port that feeds a TIG tune preselector. The preselector may be sweep tuned or manually set to a center frequency from the front panel. The output of the preselector is fed into a microwave mixer. The L.O. port of the mixer is fed with the output of an RF comb generator coving the required frequency range. Mixing of the preselected microwave signal and microwave comb signals occurs in the mixer but only one IF frequency narrow span is able to pass through the following IF filtering.  The frequency counter stages that follow calculate the original microwave signal frequency using the preselector data, that of the comb generator and the resultant IF frequency.
All oscillators are locked to a master reference so frequency errors are minimised.

This is an oversimplified decription of the design but gives the reader an idea of how the EIP 25B works in the microwave frequency ranges. The fact that the input signal is converted down to an IF frequency enables easy signal processing and the calibrated power measurement of a single carrier. This is not possible with a conventional frequency counter design or a bolometer broadband power sensor.

The EIP 25B can sweep its full Microwave capability very quickly to identify the peal signal that is found. It can also be set to sweep a narrower frequency span in order to find the strongest carrier frequency. The unit can cope with multiple carriers at the input and will select the strongest for analysis. Pretty smart and useful in the RF communications arena of operation. The calibrated level measurement capability enables the unit to measure signal level in various SI units, and RF power in mW & W.

In some ways the EIP is more of a receiver/frequency counter hybrid. In spite of higher frequency pre-scalers, the frequency selective heterodyne microwave counter is very much alive and kicking in the industry today. It offers advantages over 'dumb' frequency counters.

It is a specialise type of frequency counter however and most hobbyists would likely have little use fo rit. I work at up top 10GHz though so such a unit is very useful 

Aurora

[Lightages]

Any progress?

[pascal_sweden]

The sting in this tail of this story is that after ordering the AE20401, I saw an advert on this forum for a lovely EIP 25B 20GHz selective microwave frequency counter that includes RF power measurement above 1GHz .... perfect for my needs 

Well to cut a long story short, that 20GHz frequency counter is now on its way to me. The AE20401 has effectively been made redundant by the EIP but I am sure I will find uses for it, and it is very 'cute' 

Aurora

Those units are quite expensive! See the Ebay auction of a similar unit:
http://www.ebay.co.uk/itm/EIP-25B-Frequency-Counter-10-GHz-20-GHz-Phase-Matrix-measures-CW-FM-AM-/361149164735

Did you pay a similar amount, or did you really score a bargain deal? =)

[Fraser]

The fellow forum member gave me a great price as I also bought his VNWA3E. 20 GHz frequency counters cost thousands new and my unit is in as new condition.
My total outlay for the VNWA and EIP was 700 Euros.

The seller took account of my Retired status and was both very kind and generous in his dealings with me.

Another EIP25B sold on eBay last week for just over £300... another great deal for someone.

http://www.ebay.com/itm/Marconi-Instruments-EIP-Model-25B-Frequency-Counter-/151671564024?ssPageName=STRK%3AMEDWX%3AIT&_trksid=p2047675.l2557&nma=true&si=igOQ4x6Auz6yGDDqoMqiwWAHKig%253D&orig_cvip=true&rt=nc

Aurora

[Fraser]

@Lightages,

Sorry, I have made no progress on the build. Lots of other distractions and chores since getting the unit. I want to get on with it though so watch this space and hopefully it will be built in the next few days 

Aurora

[dadler]

They now have an up to 8GHz power meter option board:

http://www.ascel-electronic.de/kits/45/ae204014-8-ghz-power-meter-module-for-ae20401-5.8-ghz-frequency-counter-/-rf-power-meter

You can but the kit with the new board from their site, cheaper prices too.

I was hoping you would get your kit built and report back on your experience.

I might not be able to wait and may buy one myself. 

[Fraser]

It may get built tomorrow if all goes well 

Glad to hear  that they have released a new power meter module. Almost 50 Euros though. I will not need it but when buying the kit it may be the best option.

Aurora

[Fraser]

Regrettably the Ascel frequency counter build is still on the 'back burner' for the moment. I bought a 3D printer and that needed my attention. I am tied up with other duties for the next few weeks so sadly I will not be able to report on this any time soon. Sorry for any disappointment caused.

I have the new 6GHz power module on its way to me. It should be noted that the 6GHz module has a limited range of -55dBm to only -5dBm. A set of SMA microwave attenuators will be an important accessory for this module.

Aurora

[dadler]

I recommend this attenuator set:

K1-VAT2+

http://www.minicircuits.com/products/DesignerKits.shtml

Can't figure out how to link to the set directly, but it's a set of 6 2W SMA attenuators. They come in other frequency ranges and power levels, this is just the set I have and speak to.

[JohnPen]

Aurora did you get round to building this kit?  If so how well did it perform. It looks to be a an interesting build as there  are not a lot of Ghz  frequency counters in this price range.  It also fits on a small desk area better!

John

[Fraser]

Hi,

Sorry no chance to build that kit yet. Other priorities have got in the way.

I can report that the manufacturer kindly shared the full schematics of the unit and option modules with me. The module schematics are not public domain so I cannot share them here. I can say that the design is very good and I have no reason to believe that the unit will not perform well.

I do draw attention to the frequency counter specifications however. Like all microwave frequency counters you need to consider the upper and lower RF levels that are measurable. The frequency pre-scalers are not that sensitive and pre-amplifiers are sometimes needed.

The chip used in the 6GHz pre-scaler module has a stated input range of -10dBm to 0dBm. There is no additional pre-amplifier or attenuator in front of the pre-scaler IC.

The input to the pre-scaler passes through a 1nF DC blocking capacitor and a BAS40 diode limiter.

I have two of the kits and examples of both the 500MHz and 5GHz RF power meter modules.

I will report back as soon as I find the time to build these kits.

Aurora

[DimitriP]

...but if anyone wanted to roll their own,using a ZX47-40  like Jim did, here is the published article from 2011 :
http://www.minicircuits.com/pdfs/QEX.pdf

[rounin]

I just got one of these together. The build wasn't too bad, but once the LCD screen is soldered in it is really hard to rework... I think I'm going to get rid of the pins it is on and switch to a few inches of wire so the screen can be pulled off.

I did a quick test at 100 kHz and 10 Mhz, using a SDG1025.

It look to meet the +/- 1 ppm at 10 MHz, but it is a bit weird at lower frequencies. The 10 Mhz test has a nice bell curve with a std dev of ~1-2 PPM, but given a 100 kHz square wave, it jumps between 99,999Hz and 100,329Hz, basically bimodal.

Not sure if mine has assembly (or user) error, or if it is just glitchy/noisy at low frequencies. Probing the input to the counter does show a nice clean 100kHz square...

[rounin]

I used the data logger to capture my FRS-C directly for a while (no SDG1025 in the middle) with similar results:

min: 9999978.9   Hz
max:10000013   Hz
avg: 9999992.4   Hz
stdev 10.8 Hz
time   1.6 Hrs, ~5000 samples

The PC software is pretty easy to use, allows viewing the current reading and switching modes (channel A/B/C/PWR). The Pass/Fail threshold seemed to work.

Few minor bugs - the entire form updates every second which moves the text position cursor when trying to type in the min/max values for the Pass/Fail count. The graph also doesn't seem to have any zoom or pan features, but you can export the ~1 Hz samples as a csv.

I haven't installed the 5.8 GHz prescaller or power meter yet, might try that tomorrow. I don't think I will be able to generate a clock high enough to really test it though, need some more test equipment for my test equipment : ).

[TheSteve]

I haven't installed the 5.8 GHz prescaller or power meter yet, might try that tomorrow. I don't think I will be able to generate a clock high enough to really test it though, need some more test equipment for my test equipment : ).

Yes indeed, we always need more test equipment to test and verify the equipment we already have!

[bullonwheels]

I recently purchased this but getting an ~200khz on channel B if checked for frequency 2.4Ghz. Any recommendations ?

[TheBay]

Waking this thread up as there aren't any further posts on here nor other forums.

Myself and a friend ordered the AE20401 +1ppm TCXO + 5.8GHz board + 500MHz Power meter.
He initially ordered just the AE20401 + TCXO, when I ordered mine I ordered us both a 5.8GHz board and the 500MHz Power meter.

His arrived before mine so he built just the AE20401 + 1ppm TXCO kit, however using any of his references, including a 10MHz GPSD reference, a function generator and has a Gigatronics 10GHz generator the readings are all over the place, it will sometimes show the frequency but it jumps up and down by quite a few Hz so if you did not know the frequency you were inputting you would have no idea what frequency the DUT is outputting.

No issue at all with his references on a Racal Dana 1992 or other counters there.

I posted 2 videos on Youtube, is anyone else's on here behaving like this? I can only find one Youtube video with it in action reading a similar frequency and that one is all over the place.

5MHz
https://youtu.be/sCLSvnQGb5E

10MHz GPSD
https://youtu.be/VYtGDr_qtQU

[Fenstergucker]

This is a bug in the firmware. The manufacturer says it's all right, but there are some reports and measurements on the internet that prove the bug. I also have this instrument, and can confirm the error.

[TheBay]

This is a bug in the firmware. The manufacturer says it's all right, but there are some reports and measurements on the internet that prove the bug. I also have this instrument, and can confirm the error.

Yes it certainly looks like a bug in the firmware! Such a shame as it could be a really good device, I do not understand why they sell the 1ppm TCXO as it is pointless with this bug 
The reports and measurements, where on the internet is this as I'd like to read up on it.

[Fenstergucker]

I only have this link in german:
https://www.qrpforum.de/forum/index.php?thread/10258-erfahrungen-mit-ascel-ae20401/&pageNo=3

I made measurements and depending on the frequency the display always jumps around the same value. Unfortunately I don't have my notes anymore, but from 1 MHz upwards, jumps up to 40 kHz were present, always exactly the same.

I have already started to replace the microcontroller with a PIC18F with USBTMC protocol. But I haven't had time to finish everything yet.

[TheBay]

That was an interesting read!

I have been thinking about this and it has to be a firmware issue with division of binary or look up tables etc as the number is very similar every time.
Also this does not seem to happen or is as noticeable at high frequencies as the higher you go up the least significant digit changes, the 5.8GHz option is just
a prescaler, so if this was an issue with the analogue front end, calibration or source it would be even more noticeable higher up as it would be multiplied. And that's not happening.

I only have this link in german:
https://www.qrpforum.de/forum/index.php?thread/10258-erfahrungen-mit-ascel-ae20401/&pageNo=3

I made measurements and depending on the frequency the display always jumps around the same value. Unfortunately I don't have my notes anymore, but from 1 MHz upwards, jumps up to 40 kHz were present, always exactly the same.

I have already started to replace the microcontroller with a PIC18F with USBTMC protocol. But I haven't had time to finish everything yet.

[Fraser]

Maybe another email to the maker of the unit advising that the issue is now being discussed and proven on the well known and respected EEVBlog.com forum where it will be seen by many. The thought of losing sales might encourage the maker to address the issue? If not, the maker is clearly not interested in supplying a quality product and does not deserve our money. I have not used my unit enough since building it to see this problem, but I am suitably displeased to hear that the maker thought such an issue was ‘OK’.

Fraser

[TheBay]

I've sent a few e-mail back and forth with the manufacturer but they have gone quiet. They will not admit there is any issue with the Firmware  this is the last reply I had:

I've just replied to the message of your friend.
We are indeed not aware of a firmware bug causing unstable reading -
however this does occur sometimes when the input preamp adjustment is
not spot on.
One of the things we will be changing in the future is a more detailed
instruction on how to do this in the manual, and also an emphasis on how
important this is for the correct operation.
If the adjustment isn't done as it's supposed to, issues like the one
showed here may occur.

My friend had a similar reply, but was told this "It could also be a problem with the BF256 JFET, you can test this by
measuring the signal behind it - it should be as stable as the input signal.
This part is very susceptible to ESD and requires careful handling
during assembly to not damage it."

It seems they are in denial of any issues of this product and not willing to rectify it  , such a shame as it could be a great product, especially with the 5.8GHz option.
And if the source code was published then someone would have fixed this by now. I can only assume who ever wrote the original code no longer works there or something.

Maybe another email to the maker of the unit advising that the issue is now being discussed and proven on the well known and respected EEVBlog.com forum where it will be seen by many. The thought of losing sales might encourage the maker to address the issue? If not, the maker is clearly not interested in supplying a quality product and does not deserve our money. I have not used my unit enough since building it to see this problem, but I am suitably displeased to hear that the maker thought such an issue was ‘OK’.

Fraser