Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Class D Switching Power Amplifiers and Power D/A conversion

7498E BTL Mode Question
7498E BTL Mode Question
7498E BTL Mode Question Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 15th January 2020, 12:24 PM   #91
Think is online now Think  Netherlands
diyAudio Member
 
Join Date: Nov 2017
Location: Den Haag, Holland
7498E BTL Mode Question
https://www.st.com/resource/en/datasheet/tda7498e.pdf
" 220w 10%, 170w 1%@36v 3ohm."

"Overcurrent protection threshold R L = 0 Ω min 10A avg 11A max 14 A"



I don't see anything about 2ohm @ 24V mentioned in it.


If you run 2x4Ohm BTL loads, the current should be the same as with 1x2Ohm PBTL and have the same total output.

I guess the problem is that an 2 Ohm speced speaker can be lower then spec (like 1.6Ohm) just like an 4ohm speaker can be 3.2ohm.



I.e. if an 1.4Ohm driver is speced as 2ohm, it could be in trouble pulling to much power/ current. A double impedance driver, 2.8Ohm, would be speced 3Ohm, not 4Ohm.

IMO if the coil measures 2.0 Ohm (and maybe a bit lower), it should be fine to connect to TDA7498E in PBTL. The real minimum impedance is always a bit higher then the measured coil impedance and will rise if it gets warm.
  Reply With Quote
Old 15th January 2020, 03:21 PM   #92
FauxFrench is offline FauxFrench  France
diyAudio Member
 
Join Date: Dec 2017
Location: The mountains, calm and quiet.
Think, you are right. It is nowhere in the datasheet mentioned that the TDA7498E operates with a 2 Ohm load. A diligent person like you would have noticed.

My conclusion is based on logical combination of information the datasheet gives us.

I am not aware of a single class D chip that really measures the load impedance and decides to operate or not based on such information. Without post filter feedback, measuring the load impedance would be very difficult. Class D chips stops operation due to over-current, that is their detection parameter.

The datasheet, in more places, informs us that the chip can operate with two 4 Ohm loads simultaneously. Two 4 Ohm loads operated simultaneously corresponds to one 2 Ohm load if operated synchronously. It states to deliver 160W+160W (must be with 4 Ohm loads and that matches with 36V supply) but it is not clear if the chip can really do 160W+160W=320W simultaneously. You have probably seen the tiny size and in particular the very tiny leads used for the TDA7498E. As another indication that it is not 320W, the max. output power in mono configuration (PBTL) is only 220W (in 3 Ohm). Thus, there appears to be another type of limitation.

220W in 3 Ohm corresponds to 3 Ohm driven BTL from a 36V supply (the maximum supply voltage recommended in practice). 36V across 3 Ohm means 12Apeak. The not fully clear current limitation “Iocp” (table 5 of the datasheet) is between 10 and 14 Ampere (this must be momentarily, thus peak). The 12Apeak from the 220W in 3 Ohm seems to be decided by the Iocp current limit. With the tiny dimensions of the TDA7498E, this looks like a realistic limit.

With a 2 Ohm load, the Iocp limit of 12Apeak is reached with a 24V supply voltage. I see no possibility for the TDA7498E to work well with a 3 Ohm load and refuse operation with a 2 Ohm load unless the Iocp is exceeded.

My logical conclusion.
  Reply With Quote
Old 15th January 2020, 10:28 PM   #93
Think is online now Think  Netherlands
diyAudio Member
 
Join Date: Nov 2017
Location: Den Haag, Holland
7498E BTL Mode Question
That reads logical to me except " With a 2 Ohm load, the Iocp limit of 12Apeak is reached with a 24V supply voltage "



Here you mean btl (2 channels) in stead of PBTL (1 channel)?

We agree that in PBTL 2Ohm should be ok at 24V but that gives only ~125W 1% and 160w 10% THD. (about double the output of BTL@4Ohm, figure 4 datasheet) 160W/24V= 6.7A

@32V you would get ~200W 1% and ~245W 10% THD 245W/32V= 7.7A
@36V you would get ~250W 1% and ~310W 10% THD 310W/36V = 8.6A



All are within the 10A-14A I OCP (overcurrent protection threshold)


IMO, assuming 10A is the max;

A PBTL 2 Ohm load should be ok at 24V as long as it is never peaks lower then 2/(10/6.7) =~1.3Ohm.

A PBTL 2 Ohm load should be ok even at 36V as long as it is never peaks lower then 2/(10/8.6)=~1.7Ohm.


Most drivers have a peak lower then they are specified for, so I can well be that it can't drive many '2ohm' labeled speakers/drivers, but it should have no problems with a real 2.0 Ohm PBTL load.

Please correct me if I'm wrong, as this specsheet is a real 'brainbreaker' (freely translated from Dutch) for me.


(from specsheet) "160-W + 160-W output power at THD = 10% with R L = 4 Ω and V CC = 36 V"
Implies 2 channels powered at the same time.


" 1 x 220 W output power mono parallel BTL at THD = 10% with R L = 3 Ω and V CC = 36 V"
This does not rule out that it can't do 1x320W 10%@2Ohm but if it could I expect they would have mentioned it....... It just makes no sense to me.



Maybe someone mixed the 2's and 3 up?

" 1 x 330 W output power mono parallel BTL at THD = 10% with R L = 2 Ω and V CC = 36 V"



This would make some sense, but it is revision 3 and I rather expect me, then the specsheet, to be wrong.
  Reply With Quote
Old 16th January 2020, 10:58 PM   #94
FauxFrench is offline FauxFrench  France
diyAudio Member
 
Join Date: Dec 2017
Location: The mountains, calm and quiet.
Hi Think,

Both the TDA7498E datasheet and the IC itself is a “hersenkraker”.

It is a tiny IC with very tiny leads claiming to be able to handle many hundred Watts at the output. Due to the small size, surface contact with the heatsink is critical and the PCB tracks matching the very tiny IC leads become unpleasantly narrow for handling high currents. If I remember right, it was doctormord who could tell he had seen more TDA7498E ICs destroyed by over-heating. We can try to guess how the datasheet should be interpreted but for me it is a well sounding chip used at boards with an attractive price but to be doubted for high power purposes. If I needed several hundreds of Watts, I would rather go for an IRS2092, a TPA3255 or eventually a TDA8954 based amplifier.

For me the interpretation of Iocp (Overcurrent Protection Threshold), Table 5 of the datasheet, is central (but lacks clarification). The TDA7498E has separate power supply leads and ground leads for each of the two channels. That would logically mean that a current limit value could be specified for each channel. If the Iocp of 10-14Ap is such a specification, then the chip in total could handle in the order of 22A! Looking at the small IC it sounds exaggerated. If each channel could handle currents up to Iocp when connected in parallel, the resulting current limit should allow for considerably higher power than 220W in 3 Ohm. 220W in 3 Ohm is written more times in the datasheet and should at least not be a simple typing mistake.

I believe your calculation of tolerable impedance variation of a 2 Ohm speaker, assuming a 10A current limit, is not fully consistent. The 10A current limit is an immediate limit (not RMS limit) while your calculation is based on power such that the corresponding current is RMS. Evidently, I may be wrong.
"160-W + 160-W output power at THD = 10% with R L = 4 Ω and V CC = 36 V" indeed implies operation of the two channels simultaneously. With 90% efficiency of the TDA7498E, it would require 355W at the amplifier input and an RMS current of 9.9Arms. That means current peaks of 14Ap which typically exceed Iocp.
Exactly what sets the 220W/3 Ohm limit I cannot say for sure. I can see Iocp as one parameter but as you conclude, the specifications are not clear or consistent.

Why ST decided to put an apparently very power performing chip in a tiny IC housing with very tiny leads, I cannot see a logical reason for. If they expected users to take advantage of the power potential, ST could have chosen a considerably bigger IC housing such as for the STA power bridges or the TDA8954TH. The TDA7498E anyway needs a huge heatsink if used for more hundred Watts so the size of the IC would not allow a really compact construction anyway. The TDA7498E power switches have higher ON-resistance than TPA3116 claiming much less output power. It is not logical!
  Reply With Quote
Old 18th January 2020, 10:03 AM   #95
Turbowatch2 is offline Turbowatch2  Germany
diyAudio Member
 
Join Date: Aug 2013
Location: Northern germany
Hi,
you are right if you consider the TDA7498E datasheet a little bit problematic. The TDA7498 (without "E") is more complete. There you find some missing information.
There is a very short answer for these problems you discuss: The datasheet is right, your interpretation is wrong.
About your problems with calculating power, it is always misleading to apply old Ohms lawīs to D-amps that are made for playing dynamic music. In real live there is no condition of such an amp that can be described with static ohm / volt /ampere and watt.

These datasheets are not for end user, but for professionals that build end user products from them.
With such products for 99% the audible output is what counts, not what you may get feeding them for minutes with signals from a function generator.
With inefficient A/B amps such measurements like rms power made some sense, as heat sinks, capacitors and transformers had to stand the test.
With D-amps this has changed a bit. D-amps are usually feed by switching power supplies, which do not lower the voltage under load. Heat generating losses are much lower. Todays D-amps can play music near clipping level off a 1000 watt for days. If you force the same amp to deliver 1000 watt of a 100 Hz sine wave into a constant load, it will shut down after 90 seconds. Itīs constant rms power may be less than 250 watt. This is a true 1kW amp in use, but not if measured in a lab. Funny?
This comes together with the usual misconception of seeing loudspeakers as resistors. Luckily, where really much power is needed, the impedance of loudspeakers is not constant and often much higher than the static ohm value.
So if you see all of that with a basic education on electricity, you may see all kinds of problems that do not matter, while not even noticing the real ones.

So, for example, why can a D-amp play music with 500 Watt, while only having a 300 watt power supply and pulling 200 Watt from the mains. Mysterious, isīnt it?
Try to be realistic: If you think a certain amp is to weak (in reality or because of what you read from itīs data), take another one, that is stronger. Always a good way to select amps is the power rating of your loudspeaker: just take an amp that has two times the rms of the speakers and you (and your speakers) are perfectly safe.
  Reply With Quote
Old 18th January 2020, 11:59 AM   #96
FauxFrench is offline FauxFrench  France
diyAudio Member
 
Join Date: Dec 2017
Location: The mountains, calm and quiet.
Many thanks, Turbowatch2.

It makes me recall certain political statements: "The datasheet is right, your interpretation is wrong" -> "I know what I said, you just misunderstand me!"; ".....it is always misleading to apply old Ohms lawīs to D-amps that are made for playing dynamic music" -> "For such complicated matters, traditional law/economy/ethics does not apply."; "......datasheets are not for end user, but for professionals...." -> "What do you actually know about politics?!"; "......basic education on electricity, you may see all kinds of problems that do not matter....." -> "You believe to understand because you have some education but politics is art and about visions!".

OK, my personal conclusion: If you like the sound, use this chip. But, not near the limits because the limits you will only know afterward, when the chip is dead.
Actually, I like the sound.

Last edited by FauxFrench; 18th January 2020 at 12:02 PM.
  Reply With Quote
Old 19th January 2020, 06:39 AM   #97
Turbowatch2 is offline Turbowatch2  Germany
diyAudio Member
 
Join Date: Aug 2013
Location: Northern germany
Hi FF,
that is a very interesting interpretation. I like that.

My intention was to point out that with these amps things are not so easy any more. While A/B amps are easy to handle with basic Math, D-amps and SMPS messed this all up a bit.
Basically because the efficiency is so much better, our "gut feeling" we have developed with A/B amps, does not feel well with this new breed.

You can trust the data sheets, really, but you can not trust the builders of cheap China versions if they claim to have the same performance as the reverence design.

Luckily the protection circuits are included in many of the newer chips, because if they are not, some Chinese "designers" have a tendency to ignore things that they do not understand, as long as the amp plays (at least for a while). Saving the few resistors for the protection is seen as clear sign of smartness. Ask LJM, the “no schematics” specialist.

If it was not so damn cheap to produce in China, no one would buy their stuff, we have to remember this fact from time to time. If we would not care about nonsense like environment, climate change, human rights, working conditions, medical& social services and education, we could build much better amps even cheaper. Mr. Trump has realized on this and is transforming the US into China 2.0 What a smart guy. Wait until he is president for live, like Mr. Jinping!

What is really remarkable, these amp chips are so perfect, even with worst designed PCB, missing parts and cheapest materials, they sound nearly perfect. Blame that on TI, ST and International Rectifier. Really fine design!

PS today used my new router (wood work) for the second time. Has done easy work for 10 minutes now, since new. Started to sound like my old one, that was abused for 25 years, short before dying. Old one, "Metabo made in Germany", new one "GÜDE made in China".
For the old one I had to work two days, for the new one less than an hour.

Last edited by Turbowatch2; 19th January 2020 at 06:47 AM.
  Reply With Quote
Old 20th January 2020, 10:30 AM   #98
FauxFrench is offline FauxFrench  France
diyAudio Member
 
Join Date: Dec 2017
Location: The mountains, calm and quiet.
Many thanks for your response, Turbowatch2.
I had some doubt splitting facts from sarcasm.

The TDA7498E datasheet does for most chip parameters use specifications that are similar to in other datasheets such that the technical concepts are alike and the units used in accordance with their definition. The Germans were, with the DIN-norms, among the first to promote standardized concepts in order to allow unambiguous communication of characteristics.
It is mainly the TDA7498E current limit and its lacking match with specified output power levels that cause confusion. I believe it is not due to Thinks or my inferiority to professionals that there are such discrepancies.

When you hint that output power should be understood in a more dynamic context for class D amplifiers, I see this being ambiguous. The DIN-45500 norm specified continuous and simultaneous operation. This was not chosen as the best resemblance of real use, it was chosen to avoid ambiguity and to achieve comparable results. I understand deviations from this well defined but to some extent unrealistic concept as long as the deviations are clear. The crest factor effect is acknowledged but returning to dynamic power values like the “music Watts” used in the last century does not bring anything but arbitrariness, resulting in mistrust. Evidently, regulated linear power supplies that did not sag with loading existed before the SMPS became common. Undoubtedly, a TDA7498E amplifier can handle peaks equivalent to 1KW at the input but the TDA7498E amplifier will clip such peaks because Ohms law still applies. The basic principles of physics of which Ohms law form part cannot just be discarded because new principles of operation appear. I can understand that it is difficult to specify how well you in practice can cool such a small housing and that the current handling of the leads also depends on the thickness of the PCB tracks. But, I cannot understand why the datasheet only leaves us with a single “Iocp” value that is not consistent with the output power levels indicated. That is hardly because Ohms law no longer applies and not because it would be impossible to indicate in writing why there are deviations.

In my time I have worked with very bright minds. Not yet with yours. I learned that while such bright minds are able to handle impressive levels of complexity, they are not necessarily able to communicate in a clear and complete manner. As a consequence, they are not necessarily ideal for writing the datasheet while they may have added excellence in the chip.
I recently learned on the forum that many chip designs today are made by temporary design teams that are often leaving when the job is done. I can imagine the problems for the remaining marketing and customer support to draw up consistent specifications. That may be another reason for discrepancies.

Also I get astonished about certain, in more cases Asian, designs occasionally setting aside even the more basic design principles. I do not hold the TDA7498E (or any other chip) responsible for failing performance due to poor implementation. This is where we should maintain German like quality standards for our products. Else, our products have no merits.

Last edited by FauxFrench; 20th January 2020 at 10:32 AM.
  Reply With Quote
Old 20th January 2020, 05:23 PM   #99
Turbowatch2 is offline Turbowatch2  Germany
diyAudio Member
 
Join Date: Aug 2013
Location: Northern germany
Hello FF,
I think there is not much where we both disagree technically or in our view of the world, markets and manners. Also, this is not the place to look at all aspectīs that some theme, like an amp has, globally.

Ohmīs laws of course are not changed by D-amps. You just have to interpret your measurements differently. Take a 25kg Class A/B (most will be Class H in fact) and compare it to a 5kg Class D. While both will perform identical, connected to a PA and driven on a stage, the basic DIN measurements or other "old" standards will show quite different pictures.
I understand your demand for a standard measurement on one side, but also see where they do not matter in my (and your) builds, as we decide by our pick of components, what we get in the end. Usually we can predict the result with some precession.

I donīt think the data sheet problem is really existing. While these chipīs are used in millions, there is only a hand full of DIYS people that really build something with them. Iīm sure if no DIYS person world wide buys any of those, the manufacturers would not even notice, even after years.
That is what I want to say with it is not made and does not matter for us.

The professional user has unlimited access to the customer service, mind that. Second, he is not interested in real data, only the result counts and the name of the result is only sales numbers.
As you sure have noticed, the real sound quality of home audio from brands like JBL etc. (donīt say the bad word: BOSE) getīs lower with any generation of new gadgets. HIFI has not evolved noticeable in the last years, in fact it is in a stage of devolution.
Most of the friends of my teen/ twen daughters have no idea how realistic and impressive music reproduction in a living room can be.
They are raving how well ALEXA sounds. If you compare these data to noise converters with even a 1990 Denon/ Onkyo/ Sony mini Hifi set, be prepared to instantly vomit.

The industry has realized what itīs worst enemies are: Quality, product life, environment, ecology. The best productīs have integrated batteries, so they 100% fail after a limited time. Compare that to your 1976 HIFI amp...
Quarterly reportīs matter, not climate change!
  Reply With Quote
Old 21st January 2020, 10:07 PM   #100
FauxFrench is offline FauxFrench  France
diyAudio Member
 
Join Date: Dec 2017
Location: The mountains, calm and quiet.
Let's leave the discussion like that, Turbowatch2.
A general concern I have is that not only the technical quality of many products degrade, also the documentation becomes in many cases ridiculously poor. Sales figures rule and both sellers and buyers know less about what is being sold.
"Think", we will most likely never know the exact truth about the current limit(s) of the TDA7498E. I am pretty sure it can PBTL-handle 2 Ohm loads to a certain level and when it stops playing, we know we have passed the limit.
  Reply With Quote

Reply


7498E BTL Mode QuestionHide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
TA2022 vs 7498E digital T-amps ljbrandt Class D 23 14th May 2018 04:34 AM
7498E and AptX DDF Class D 1 10th May 2017 12:54 PM
Pentode mode, second B+ question riven67 Tubelab 2 11th October 2016 09:11 PM
Switch Mode PS question (Chip Amp) Tall Shadow Power Supplies 7 30th October 2007 05:17 AM
Question about Mode 11 of CS8412 quantran Digital Source 2 17th April 2007 12:58 AM


New To Site? Need Help?

All times are GMT. The time now is 12:26 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2020 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2020 DragonByte Technologies Ltd.
Copyright ©1999-2020 diyAudio
Wiki