Synthesized/Arbitrary Waveform/Function Generators: sampling rate matters

It’s basic (signal processing) mathematics that square waves (or any waveform with sharp edges) carry harmonics that doesn’t die fast enough, therefore not band-limited. The more sudden the transition is, the more harmonics you need to preserve to faithfully represent a signal in practice.

Unless the signal itself is known to be band-limited (like sine waves, classical modulation schemes, SRRC), it puts the burden on the test instruments involved to provide large bandwidths and the matching high sampling rate (needs to do better than Nyquist).

The technology today provides generous bandwidths and sampling rates for oscilloscopes at reasonable prices, but synthesized function/waveform generators with the same bandwidth/sampling rate can easily cost 10 times if not even more. For the money to buy a used not-too-old 500Mhz pulse generator, you can buy a 4GHz 20GS/s oscilloscope!

For oscilloscopes, users are often aware that the rounded square waves/transitions they see on the screen is due to bandwidth limitations, and will account for the reality distortion in their mind. If your square wave clocks are pushing the sampling rate of the scope and the combined bandwidth between the scope and the probes, you should very well expect that you cannot catch much glitches and pretty much use the oscilloscope as a frequency counter or check for modulations.

For synthesized signal generators, bandwidth and samplers are way more precious. Oscilloscopes generally has 8-bit vertical resolution (256 steps), but synthesizers typically has 12-bit vertical resolution (4096 steps) or better. There are imperfect techniques trading vertical resolution and sampling rate, but there is no free lunch. 

I have on my bench these 12-bit function generators:

  • an old Analogic 2045B (roughly US$1300, 400Mhz, 800MS/s, 2Mpts) and
  • a much more modern Agilent/HP 33120A (roughly US$600, 15Mhz, 40MS/s, 16Kpts)

that I’d like to illustrate the value of getting a higher bandwidth (and therefore higher sampling rate) unit. The 33120A has a much finer control over frequency/amplitude/offset steps (2045B only allows fixed point increments) and might have better noise characteristics and much smaller form factor considering Analogic 2045B is made in the 1970s and HP 33120A is made at least 20 years after that. I would have liked to keep only my 33120A or 33255B on my bench to save space, but once you’ve seen this screenshot, you’ll know why I’m still willing to cough up the space for 2045B:

The upper waveform (Channel 1, yellow) is a from Analogic 2045B while the lower waveform (Channel 2, green) is from HP 33120A. They are both set at around 15Mhz and you can see that when approaching the limit of the synthesizer, 33120A rounds off the square wave to almost a sinusoid.  Square waves gets ugly quickly above 10Mhz and this is as far as 33120A is capable of.

On the other hand, the square waves are bang-on for 2045B, and is still decent at around 50Mhz (my BNC cable starts to come in question). That’s why it’s still worth getting a high bandwidth synthesized function generator even if your budget only allows for clumsy old models if you use your function generator for something more than sinusoids.


Note that function generators are typically designed to pump out to 50Ω loads and the amplitude displayed in the function generator (V_{gen}) assumed so. That’s why beginners gets confused why they read 2V_{pp} when the function generators says 1V_{pp}: the function generator sends out nearly double the voltage so the potential divider formed between the 50Ω output impedance of the function generator and the 50Ω load impedance will split the voltage into half. If you set your oscilloscope to 1MΩ load, you are getting (2V_{gen})\frac{1MΩ}{1MΩ+50Ω}, which is nearly 2V_{gen}.

More importantly, if the load is not matched, the small capacitances in the chain will distort the waveform received by your oscilloscope severely at higher frequencies, so you can barely get a square wave at fundamental frequency above 2Mhz undistorted if you feed it into an oscilloscope with 1MΩ input impedance, while in reality the signal generator, cables, and oscilloscope can do much better.

Most cheap low bandwidth oscilloscopes (like 100Mhz) do not have 50Ω option. Nonetheless the impedance needs to be matched if you work with square waves at 2Mhz or above. Just buy a 50Ω feed-through BNC ‘termination’ adapter and plug it right at the 1MΩ input port. In reality, it’s a divider between 50Ω and (1MΩ//50Ω), with the scope seeing \frac{(1M//50)}{(1M//50)+50} of 2V_{gen}. For all practical purposes the oscilloscope sees (2V_{gen})\times0.5 or V_{gen}.

With 50Ω feed-through BNC ‘termination’ adapters, make sure you work out the impedance matching if you split the signals if it’s not the simple nearly 1:1 potential divider halving the voltage. At low frequencies, the amplitudes will be off, but when you start going into Mhz range and above, your signal will be distorted badly as well.

 

25 total views, 4 views today

 

25 total views, 4 views today

Introduction to funny colloquial HK Cantonese phrases: 搵食啫,犯法呀 That's how we justify outrageous things we do that's still technically legal

This summarizes the cultural mentality in HK:

  • as far as profit is concerned, anything goes as long as it’s legal
  • schadenfreude

The third quote in the video was actually taken out of context and therefore incorrectly translated. In the show, the word 精神 refers to (body) energy, not spirits (ideals). Dayo meant to say that HKers so hard in a prospering economy that people were constantly tired and lacked the attention span to appreciate art and culture. Just ignore it.

Nonetheless, I believe Mr. Siu made the mistake because we all subconsciously agree that the core values of Hongkongers is the lack of thereof, which the stand up comedian Dayo Wong also mentioned in his (supposedly) last standup comedy show. In fact, I’m proud of it: it simply means even the average HK people aren’t too stupid (see Dilbert and Rick and Morty and ye shall comprehend: values are for stupid people who cannot reason through the purpose of their actions)

28 total views, no views today

Everything is conditional. We just don’t know what the conditions are.

This is a quote from House MD. It succinctly explained my entire viewpoint of everything so far with un-contradicted accuracy.

With this concise wording, I don’t have to explain to people why I don’t want any romantic relationships, why I think the concept of family is toxic, and why I consider love as a vestigial trait from evolution that we’re are abusing to mislead people to do our bidding and not pay them fairly. Now they can work out these conclusions by simply reasoning from “Everything is conditional. We just don’t know what the conditions are .” Neat!

17 total views, 2 views today

TDS 500~700 series Power Supply Recapping A16 (620–0063–04), also Zytec 22917401

While servicing a TDS 754A for a client, I smelled burnt electrolyte near the power supply section. Although it isn’t the cause of the problem yet, I know it’s a ticking time bomb.

By comparing the good power supply from my TDS 784A (same base design), I saw one of the leads of a higher power diode looks corroded (black stuff) yet the same diode on the good power supply has rainbow discoloration. It suggested that the assembly of the same part number is likely to fail by poor design (must be heating too close to the capacitor). Here’s the comparison of the C49 that caught my attention:

And after removing C49 on TDS 754A, it’s clearly this capacitor has leaked and corroded one of the diode’s lead nearby:

By taking a closer look, I noticed a bit of stains around most 2700uF 10V Nichicon capacitors. Only C86 and C30 haven’t leaked yet. Might as well replace them all since there are 8 of them and 6 of them leaked.

C85 has green stuff all over it and smelled horrible. Surprisingly the ESR and capacitance is still within specs. That’s why the unit still functions. It’s just a matter of time before the power supply blows up and take out the commonly known transistors with it if I had left it there:

C47 and C48 is a mess:

C43 doesn’t look too bad, but it actually leaked. The clear fluid there is not flux, and the diode leads nearby stained for a reason:

C29 and C26 leaked as well:

C30 near them is clean though, the one out of two survivors:

Despite I haven’t seen leak residue on the PCB for the 680uF 35V, they are located close to high heat areas so I desoldered them to take a look. Turns out C21 cracked,

C44 leaked a little, C42 is intact (it’s just flux):

All 680uF there are Marron capacitors.

So basically, I couldn’t trust the caps anymore and I desoldered the rest to check for leaks. Some of the Matsushita / Panasonic branded tinier capacitors far from heat sources survived. The 100uF 25V capacitor (Matsushita) at C33 near the heatsink also leaked, but it’s not too visible until I see the corroded pads after desoldering it.

I took out the last Nichicon there, a smaller 47uF 80V at C17, despite I don’t see any visible leaks before I desolder it. Glad that I did. It clearly leaked (can see it by looking at the bottom of the extracted capacitor), but not outside the capacitor’s casing’s diameter:

To avoid troubleshooting nightmare (uncommon problems) in the future, replace ALL electrolytics on the power board regardless of whether they are good or not given the majority of the capacitors leaked in this example.  If you leave one or two old capacitors there and they leaked in the future, it’d be an uncommon problem that you can’t get any advice anywhere since nobody serviced the unit the same way as you did.

To be fair, Tektronix didn’t make this 400W power supply, Zytec did:

 


I used to think that the TDS 700 series doesn’t need much work because the SMD aluminum electrolytic capacitors on the acquisition board. But now I can see that anything that’s electrolytic leaks (CRT driver, power supplies, front-panel keypad, RS-232/Centronics board, processor board) in this TDS 500~700 series.

Nonetheless, it’s still a positive trait that there are no electrolytics on the TDS 700 series acquisition board, as it’s the most expensive and fragile piece.  Acquisition board with leaked electrolytes is toasted (beyond economic repair) if you leave the electrolyte there too long.


Do NOT buy TDS 300~800 series off used market if you do not have to (like you have automation written for it or you’ve used it for 20 years and it’s all ingrained in your head) no matter how cheap they are (or SEEMINGLY working). The money is much better spent on HP 54520/54540 series if you are on a very tight budget. TDS 300~700 series don’t have much usable life left unless it’s verified new-old-stock. All fixes to TDS 300~700 problems are are laborious, frustrating and expensive.

It’s the same things that breaks for the same reason (unreliable design). That means if you simply swap modules with another used unit, or buy another identical unit, you are going to run into problems one way or the other in a short amount of time.  Basically, you are only squeezing the last few puffs off a disposed cigarette butt.

I have built the knowledge and parts to rebuild these congenitally sick puppies, but as I discovered the number of common problems are still growing strong, I’m staying out of the market for it and sell whatever I have left (I’ll strengthen them before selling, of course).


If you absolutely have to rebuild a TDS 300~800 series oscilloscope and are willing to spend good money on it, which is typically the case if you:

  • have an automated system written for it that you need an exact replacement
  • have used the unit for 20+ years that you’d willing to pay to not painfully relearn.
  • do not want to change the procedures in a bureaucratic environment

I have the parts and knowledge to extend the unit’s life that you cannot find anywhere else. It’s super involved, but I’d be willing to help if I’m the last resort.

If you choose to send me a unit for rebuild to extend its life, I’ll make it mandatory to replace electrolytics capacitors in these boards:

  • Processor board
  • RS-232/Centronics board (Option 2C)
  • Front panel keypad
  • CRT driver board
  • Power supply module
  • Acquisition board (if your model uses SMD aluminum electrolytics).
    Acquisition board cost a lot more to recap as there’s a lot of capacitors if the model uses any.

because electrolytic capacitor failures cause symptoms that are very hard to troubleshoot (most of those are power rail capacitors, which if they fail, unstable voltages gives unpredictable erratic behavior).

The following is optional and billed separately:

  • New CRT tube for color CRT screens. I have six units left so far. First come, first served.
    Tuning the tube to match the CRT board is very labor intensive.
  • Rebuild attenuator hybrid (they are consumables)
  • Troubleshoot/repair existing known symptoms

I give 3 years warranty for the repairs or preventative service I’ve carried out and it’s not user inflicted damage after the repair (like feeding high voltage to the inputs).

Call me at 949-682-8145 if you are truly need to rebuild a TDS 500~700 model and is willing to pay good money for it.

46 total views, 1 views today

TDS 500~700 base designs

TDS 500~700 series uses common base design depending on when is the time range the model is produced, so the model number itself doesn’t tell you much about commonalities. For example, TDS 520 is common with 540, 620, 640 because they are all the first generation produced by SONY. Their main PCBs assemblies are significantly different from later ones like TDS 540A (Note the ‘A’). They don’t even use NVRAM chips with the same pin-out.

Yet TDS 540B is very different from 540A as it has InstaVu and no SMD aluminum electrolytic capacitors. It’s another generation. Yet even more confusing is that ‘A’ and ‘B’ does not represent different generations across the board. It only ties to the generation associated with the base model number. For example, TDS 500B, 600B and 700A has the same basis (and therefore the same service manual).

So far, service manual is the sure-fire way to tell what models shares the same design. They only removed a few components and ID resistors to make a lower-end version for market differentiation. The prices are no longer consistent in the used market, so sometimes it might be possible just to takes parts from a higher end unit and downgrade it with resistor ID for repairs. TDS boards is ‘adjusted’ before they ship, and has more mechanisms (like bandwidth-limiting resistors), so a lot more work is involved if you want to move up. I heard from forums that if you try to turn a monochrome processor board into color processor board, you’ll have to install extra chips and components.

 

17 total views, no views today

Agilent HP 54810-66501 didn’t even trim the leads for the 100Mhz crystal 54810-66501 is the same base PCB for 54810-66527 54815-66527 54820-66527 54825-66527

Louis Rossmann tored a fake Hakko soldering station down and was stunned to see the IC leads not trimmed, a clear sign of lousy manufacturing.

I noticed the long pins of through-hole a crystal oscillator on a 54810-66501 acquisition board, coming from a well-made Agilent/HP 54810A/54815A/54820A/54825A oscilloscope (I know people complained about these oscilloscopes, but most of the failures is in the computer section, not on the acquisition board side. I know the computer section very well, so no problem for me.)

 

24 total views, no views today

Is it worth repairing the Agilent/HP Attenuator Hybrid (1NB7-8303)?

I have a few worn attenuators and one that I received that was fried by high voltage and I tried to swap the relays. Turns out it’s not really about swapping the coil, but a near impossible precision task if you want to swap the entire block without opening up the contacts and magnet gliders. If you desolder the coil pins, you can release them and expose the inner workings:

Usually the relay coil is not the problem. It’s either the magnetic shuttle (the black stuff between the two coils) that’s not moving smoothly or the contact metal spring does not naturally bend to make good contact anymore. I fixed the first one with WD-40 (the magnet glides on a custom plastic rail), so some vertical divisions that used to be capacitively coupled (i.e. there’s an air gap instead of good electrical contact) were fixed, but it still won’t pass calibration because of the worn metal spring. Here is what the spring(s) looks like:

To put the motor coils back, I slightly push it down to the board while guiding the shuttle (that has a tiny piece of magnet in it) with a strong magnet outside the coil housing. It will fall in place easily.

Given how reasonable watronics (Bill Watry) is charging for the attenuators, it’s not worth the time, effort, and uncertainty trying to perform the surgery. He basically serves any HP/Agilent instruments that uses this attenuator hybrid that looks like this.

Bill Watry is a veteran of the 54500 series, which is the main consumer of this kind of hybrids. He’s the first person to talk to if you have any problem with HP 54500 series oscilloscopes. Please contact him directly rather than through eBay if you can, as eBay charges hefty fees (it eats up 13% of the transaction amount, not what he earned after costs).

54610B/54615B/54616B/54616C as well as first generation Infiniium uses this kind of attenuator too. I have everything needed to service 54615B/54616B/54616C except attenuators. If it boils down to attenuators, I don’t stock them and you’ll have to order it from Bill (I can do that on your behalf if I’m the one doing the repairs).

If you have an HP Infinium or Agilent Infiniium and your situation likely involves the computer section, I should be the first person to talk to, since I got nearly all the nasty quirks down over the last decade so you don’t have to spend months navigating through this minefield. The learning curve is really steep if anybody tries to figure it out on their own for the first time.

37 total views, no views today