Strobe & Flash Output Data
Moderators: rjlittlefield, ChrisR, Chris S., Pau
Thanks for the clarification and info on the tubes. With similar gas mixture tubes about the same size, then you might expect the output power "waveform" to be similar for the same cap & voltage.mjkzz wrote:I tend to disagree with this.Point being, if two strobes/flashes have the same energy storage capacitance, same storage voltage they probably have similar optical outputs.
I built a 400Ws IGBT flash myself as DIY project (one year before any Chinese 200ws range products, just to brag) and I have experimented with different flash tubes. Some tubes discharge much faster than others, so you will see shorter tailed graph and they perform much better in terms of consistency of color temp, etc.
The speedlites you are working with probably have same kind of gas mixture inside flash tubes, so they have same characteristics if voltage and capacitance used are the same.
Xeon tube for police/emergency vehicles discharge much slower than those used in photography, I once had a such tube that took a whopping 12ms to discharge same amount of energy as a tube used in Canon 580EX which took about 4-5ms. This is actually good for high speed sync (or pseudo HSS), but that another story.
The total delivered optical power is the integral of the waveform, and this should depend on many things. The efficiency of the tube in converting current to photons and the drive electronics (cap, switch, wiring and so on). All of which I would think should influence the total optical powered delivered. The waveforms may be different but the power may be similar.
That makes sense regarding the color temp which might be related to the actual gas temperature when ionized. A slow decay seems it might "smear" the color as the gas cools from the initial hot temp at initial ionization.
You bring up an interesting point which seems the optical tube efficiency could be estimated in simply measuring the tube temperature after the flash. Of course many things influence this, but a relative measure would be interesting. I have a cheap IR temp sensor I might try and see. I know the tubes get pretty hot after a full power burst, so lots of energy is being converted to heat and not photons!!
Thanks for clarification and information. BTW what size caps and storage voltage did you use in your DIY project?
Edit. I just triggered the Interfit EXT400 strobe 10 times in a row at full power and saw the tube temperature rise from 79 to 178 degrees. The Yongnuo YN568EX speed light triggered 10 times at full power and the front plastic diffuser plate (can't get to the tube) temperature went from 77 to 97 degrees, and the Nikon SB-800 front plate reached 95 degrees.
Cheers,
Mike
Last edited by mawyatt on Sat Jan 16, 2016 11:07 am, edited 1 time in total.
Very interesting, thanks Chris. Have you seen the guy that built the massive capacitor array (filled an entire room, well not exactly but BIG capacitors) to discharge into a ultra high field electromagnet? This would shrink a quarter into a dime size, but still retain the surface details!ChrisR wrote:Remember Stephen Dalton , many years ago now, making very short duration flash? Special tubes, very special capacitors, special Ice Cream containers to hold them in?
He had trouble getting the big fast capacitors because the military used them for triggering things.
http://www.capturedlightning.com/frames/shrinker.html
Best
Mike
I had another thought (dangerous). Power IS proportional to V², so presumably the power in the flash is represented by the area under a different curve - of V²?
Or is that wrong too?
It was in part prompted by those last two scope traces. The area under the half-power trace looks too small , but if the vertical axis were squared it might look right?
[Edit - not vertical axis squared, but vertical values squared...]
Or is that wrong too?
It was in part prompted by those last two scope traces. The area under the half-power trace looks too small , but if the vertical axis were squared it might look right?
[Edit - not vertical axis squared, but vertical values squared...]
Last edited by ChrisR on Sun Jan 17, 2016 1:53 am, edited 1 time in total.
Chris R
Chris,ChrisR wrote:I had another thought (dangerous). Power IS proportional to V², so presumably the power in the flash is represented by the area under a different curve - of V²?
Or is that wrong too?
It was in part prompted by those last two scope traces. The area under the half-power trace looks too small , but if the vertical axis were squared it might look right?
Interesting thoughts for sure. I honestly haven't thought about it enough to be absolutely sure. The graphs I am about to post seem to agree with you regarding the vertical axis (squared), at least at the lower power ends where the waveforms are similar.
Take a look and see what you think. The graphs for Power setting of 13, 25 & 50 seem to follow what you are saying. This assumes the Power Setting on the EXT400 is somewhat accurate as ratios go.
I have to get some other work done (cleaning, which I know how to do:roll: , so I'll be gone for awhile.
Cheers,
Mike
Last edited by mawyatt on Sat Jan 16, 2016 11:30 am, edited 2 times in total.
@mawyatt
I used many different capacitors, for the 400ws experiment, I actually used 8 1500uF 330V (charged to 320V) caps salvaged from 20 years old Bowen strobes. These British made caps are better than the Chinese made ones, even though they were 20 years old.
You are absolutely right, many things can influence performance. Low quality caps, for example, heats up significantly more than better caps -- due to internal resistance and even internal inductance (the way they wire things inside cap). Imagine, current going through circuit can reach over HUNDREDS of amps at peak, that does not bold well for cheap components.
@ChrisR : I think you are right, it is easy to verify if scope data can be loaded into computer. Integration of curve (or at least proportionally) is really easy, if sampling frequency is constant, then integration is just the sum of all sample points. Then we calculate a straight sum and a squared sum for both full power and half power. If sampling frequency is high enough (ie, dt is small enough), this should approximate the integral.
I used many different capacitors, for the 400ws experiment, I actually used 8 1500uF 330V (charged to 320V) caps salvaged from 20 years old Bowen strobes. These British made caps are better than the Chinese made ones, even though they were 20 years old.
You are absolutely right, many things can influence performance. Low quality caps, for example, heats up significantly more than better caps -- due to internal resistance and even internal inductance (the way they wire things inside cap). Imagine, current going through circuit can reach over HUNDREDS of amps at peak, that does not bold well for cheap components.
@ChrisR : I think you are right, it is easy to verify if scope data can be loaded into computer. Integration of curve (or at least proportionally) is really easy, if sampling frequency is constant, then integration is just the sum of all sample points. Then we calculate a straight sum and a squared sum for both full power and half power. If sampling frequency is high enough (ie, dt is small enough), this should approximate the integral.
I have already replaced the voltage doubling cap in two of the cheap C-180 Neewer strobes, I had another C-180 fail and get replaced under warranty. This cap exploded in both units, due to overheating (I*I*ESR) causing overpressure and rupture.mjkzz wrote:@mawyatt
I used many different capacitors, for the 400ws experiment, I actually used 8 1500uF 330V (charged to 320V) caps salvaged from 20 years old Bowen strobes. These British made caps are better than the Chinese made ones, even though they were 20 years old.
You are absolutely right, many things can influence performance. Low quality caps, for example, heats up significantly more than better caps -- due to internal resistance and even internal inductance (the way they wire things inside cap). Imagine, current going through circuit can reach over HUNDREDS of amps at peak, that does not bold well for cheap components.
@ChrisR : I think you are right, it is easy to verify if scope data can be loaded into computer. Integration of curve (or at least proportionally) is really easy, if sampling frequency is constant, then integration is just the sum of all sample points. Then we calculate a straight sum and a squared sum for both full power and half power. If sampling frequency is high enough (ie, dt is small enough), this should approximate the integral.
I must admit I am abusing all my strobes with stacking sessions running up to 600 individual images separated by only 10~15 seconds between flash triggers. The replaced caps are working OK so far. The Interfit and Impact strobes have held up well without any issues.
100s of amps running around during tube ionization surely stresses things and exposes any weakness.
Best,
Mike
Wow, Neewer people should put a thermal circuit in side their strobe. Sure it is a little more expensive, but certain cheaper than replacing broken unit due to overheat and damage their rep.
From last few scope captures of EXT400, it seems that the unit is NOT an IGBT unit. Older strobes set power output level by limiting voltage charged on caps. I would expect the lower power level graph to have longer tail, but your graphs does not show significant long tail.
I guess having a high end scope helps vs my DIY microphone based scope :-) Thanks for the graphs, I now learned more.
PS. Nice illustration of RF signal and actual triggering. See, a lot of people set their shutter speed at 1/200 or 5ms -- shutter opens for 5ms. But because of that RF delay, actual exposure (by Flash) is only 5-1.4ms=3.6ms. If speedlite is set to full power and flash duration is 5ms, then the image will not be exposed as same as without the RF. Just a thought.
From last few scope captures of EXT400, it seems that the unit is NOT an IGBT unit. Older strobes set power output level by limiting voltage charged on caps. I would expect the lower power level graph to have longer tail, but your graphs does not show significant long tail.
I guess having a high end scope helps vs my DIY microphone based scope :-) Thanks for the graphs, I now learned more.
PS. Nice illustration of RF signal and actual triggering. See, a lot of people set their shutter speed at 1/200 or 5ms -- shutter opens for 5ms. But because of that RF delay, actual exposure (by Flash) is only 5-1.4ms=3.6ms. If speedlite is set to full power and flash duration is 5ms, then the image will not be exposed as same as without the RF. Just a thought.
I don't know if this was a design flaw (overstressed cap), or a flawed cap (bad parts from supplier).
Agree, the strobes do not seem to utilize a series switch (IGBT or BJT) with the tube so they can control the power by controlling the optical pulse width. The strobes seem to use a voltage doubler to create ~330vdc across the energy storage cap, but I don't know if they use an adjustable voltage regulator to set the storage voltage.
Actually now that I think about it if you used a SCR or TRIAC in the voltage doubler you could easily control the storage voltage by controlling the "ON" time of the SCR or TRIAC. This is probably how it's done, but honestly don't know.
You are way more familiar with these circuits than I, maybe you could elaborate on this....I am just speculating
I recall the Neewer RF trigger is the cheap type that operates in the 433MHz band (same as your car keyfob). Others operate at 2.4GHz (Yongnuo for example) and may have a different delay, something I will look into, given time and $.
Anyone care to donate?
Edit: You bring up a good point about actual camera sensor exposure time vs. shutter, strobe, and RF delay times. Another possible source of having the exposure not being what you (or camera) wanted!!
Cheers,
Mike
Agree, the strobes do not seem to utilize a series switch (IGBT or BJT) with the tube so they can control the power by controlling the optical pulse width. The strobes seem to use a voltage doubler to create ~330vdc across the energy storage cap, but I don't know if they use an adjustable voltage regulator to set the storage voltage.
Actually now that I think about it if you used a SCR or TRIAC in the voltage doubler you could easily control the storage voltage by controlling the "ON" time of the SCR or TRIAC. This is probably how it's done, but honestly don't know.
You are way more familiar with these circuits than I, maybe you could elaborate on this....I am just speculating
I recall the Neewer RF trigger is the cheap type that operates in the 433MHz band (same as your car keyfob). Others operate at 2.4GHz (Yongnuo for example) and may have a different delay, something I will look into, given time and $.
Anyone care to donate?
Edit: You bring up a good point about actual camera sensor exposure time vs. shutter, strobe, and RF delay times. Another possible source of having the exposure not being what you (or camera) wanted!!
Cheers,
Mike
I just completed testing another (Viltrox JY-02) RF trigger that operates at 433MHz, it's delay was similar to the Neewer one displayed above.mawyatt wrote:I found that there is quite a delay from when the camera initiates a strobe trigger thru an RF trigger (Neewer) to when the actual optical output happens. I have measured around 1.2~1.4ms for various strobes. Almost all the delay is in the RF trigger. See the graph below.
Cheers,
Mike
Mike
A Yongnuo YN560 IV Flash mounted to a YN622N 2.4GHz RF Trigger was used to test the delay of the YN622N Triggers. I utilized another YN622N to Trigger the YN622 with the YN560 mounted to it on the hot shoe. So basically one YN622 triggers another Yn622 which triggers the YN560 thru the hot shoe.
The YN622 delay (~500us) is less than the 433MHz RF Trigger shown before (1200~1400us). These would be a better RF trigger system for high speed work.
Best,
Mike
The YN622 delay (~500us) is less than the 433MHz RF Trigger shown before (1200~1400us). These would be a better RF trigger system for high speed work.
Best,
Mike
If my experience with Yonguo 602 RF triggers on Nikon speedlights is any guide, the effect of this propagation delay will be a solid black band at the bottom of some images. (Or perhaps the top, if the camera's rolling shutter goes the opposite direction from that of my Nikon DSLRs.)mawyatt wrote:You bring up a good point about actual camera sensor exposure time vs. shutter, strobe, and RF delay times. Another possible source of having the exposure not being what you (or camera) wanted!!
I saw this intermittently, about one shot in thirty or forty. This was with shutter speed set at 1/200 sec, already reduced from the sync speed of 1/250 sec for increased fault tolerance. Slowing the shutter speed a bit more, to 1/160, fixed the problem.
--Chris